Mgmt 587b Customer Value Management Case
MGMT 587B: Customer Value Management
Subimal Chatterjee
Professional MBA Program (New York City)
Sept. 13 and 20, 2014
Customer Value Management: Readings and Cases
MGMT 587B: Customer Value Management
Binghamton University
THIS PRINT COURSEPACK AND ITS ELECTRONIC COUNTERPART (IF ANY) ARE INTENDED SOLELY FOR THE PERSONAL USE
OF PURCHASER. ALL OTHER USE IS STRICTLY PROHIBITED.
ISBN 1269608886
XanEdu™ publications may contain copyrighted materials of XanEdu, Inc. and/or its licensors.
The original copyright holders retain sole ownership of their materials. Copyright permissions from third parties have been granted for materials for this publication only. Further reproduction and distribution of the materials contained herein is prohibited.
WARNING: …show more content…
COPYRIGHT INFRINGEMENT IS AGAINST THE LAW AND WILL RESULT IN
PROSECUTION TO THE FULLEST EXTENT OF THE LAW.
THIS COURSE PACK CANNOT BE RESOLD, COPIED
OR OTHERWISE REPRODUCED.
XanEdu Publishing, Inc. does not exert editorial control over materials that are included in this course pack. The user hereby releases XanEdu Publishing, Inc. from any and all liability for any claims or damages, which result from any use or exposure to the materials of this course pack.
− In print copy only
− Online only
Unmarked items are available both in print and online
Customer Value Management: Readings and Cases
Table of Contents
“Setting Value, Not Price” by Leszinski, Ralf; Marn, Michael V
1
“Apollo Hospitals: Differentiation through Hospitality” by Kulkarni, Suhruta;
Makhija, Kripa; Kumar, Unnikrishnan Dinesh
19
“Soren Chemical: Why is the New Swimming Pool Product Sinking? (Brief
Case)” by Rangan, V. Kasturi; Yong, Sunru
39
“Note on Behavioral Pricing” by Gourville, John
47
“A Practical Guide to Conjoint Analysis” by Wilcox, Ronald T.
59
“Fidelity Incorporated: Pricing the Fidelity Blue Chip Growth Fund” by Wilcox,
Ronald T.
69
“Portland Trail Blazers” by Wilcox, Ronald T.
75
XanEdu Extra
An Excel−formatted spreadsheet containing the exhibits for the case above is available at http://content.xanedu.com/ds/m−0773x.xlsx “Customer Profitability and Lifetime Value” by Ofek, Elie
87
XanEdu Extra
An Excel−formatted spreadsheet containing the exhibits for the case above is available at http://content.xanedu.com/hs/503019p2.xls “The Brita Products Co.” by Deighton, John
97
XanEdu Extra
An Excel−formatted spreadsheet containing the exhibits for the case above is available at http://content.xanedu.com/hs/500024p2.xls “The Independent Adviser for Vanguard Investors” by Pfeifer, Phillip E.
115
125
Bibliography
i
ii
PRICING
Setting value, not price
© THE CURTIS PUBLISHING COMPANY
98
THE McKINSEY QUARTERLY 1997 NUMBER 1
1
279
The first task is to map benefits versus price – as the customer sees them
Bear in mind that equal value doesn’t mean equal market share
The key decision: do you stay on the line of value equivalence, or get oƒf ?
Ralf Leszinski • Michael V. Marn
A
MANUFACTURER of high-quality medical testing equipment introduces a vastly improved version of its bestselling diagnostic device at a price 5 percent higher than that of the older model it replaces. For three months, the new model is successful, gaining rave reviews from customers and increased market share. One month later, prices in the sector collapse and the company has to discount its superior new product just to maintain its traditional market share.
A highly regarded manufacturer of commercial paper prides itself on delivering extremely consistent quality and service. That consistency notwithstanding, the company is baƒfled by vacillations in its market share that accompany shiƒts from tight to loose supply in the industry.
A consumer packaged goods company executes one of the most common business tactics – it matches a competitor’s price on a large contract to supply a leading food retailer. In the months that follow, a bitter price war breaks out, destroying almost all of the industry’s profitability in this product category.
These disparate cases have at least one thing in common: apparently sound marketing strategies and tactics that produced unexpected and costly results.
But could they have been avoided? Here we will explore how these and other common and expensive marketing missteps might be averted by applying a discipline called “dynamic value management” to the pricing and product positioning that are at the core of what most marketers do.
“Value” may be one of the most overused and misused terms in marketing and pricing today. “Value pricing” is too oƒten misused as a synonym for low
Ralf Leszinski and Mike Marn are principals in McKinsey’s Atlanta and Cleveland oƒfices, respectively. Copyright © 1997 McKinsey & Company. All rights reserved.
THE McKINSEY QUARTERLY 1997 NUMBER 1
2
280
99
SETTING VALUE, NOT PRICE
price or bundled price. The real essence of value revolves around the tradeoƒf between the benefits a customer receives from a product and the price he or she pays for it.
The management of this tradeoƒf between benefits and price has long been recognized as a critical marketing mix component. Marketers implicitly address it when they talk about positioning their product vis-à-vis competitors’ oƒferings and setting the right price premium over, or discount under, them. Marketers frequently err along the two dimensions of value management, however. First, they fail to invest adequately to determine what the “static” positioning for their products on a price/benefit basis against competitors should be. Second, even when this is well understood, they ignore the “dynamic” eƒfect of their price/benefit positioning – the reactions triggered among competitors and customers, and the eƒfect on total industry profitability and on the transfer of surplus between suppliers and customers.
To illuminate the nature and magnitude of this missed value-management opportunity, value needs to be defined properly. Customers do not buy solely on low price. They buy according to customer value, that is, the diƒference between the benefits a company gives customers and the price it charges.
More precisely, customer value equals customer-perceived benefits minus customer-perceived price. So, the higher the
Exhibit 1 perceived benefit and/or the lower the price
Value map of a product, the higher the customer value
Stable market shares and the greater the likelihood that customers will choose that product. (We will return to this later.)
Perceived price
Static value management
Value equivalence line (VEL)
Many marketing and strategic assessments can be made by using a simple tool called a value map, and by considering how customers are distributed within the map for a given segment.
The value map explores the way customer value and the price/benefit tradeoƒf work in real markets for a given segment (Exhibit 1).
Customer value = perceived benefits minus perceived price
The horizontal axis quantifies benefits as perceived by the customer; the vertical axis shows perceived price. Each dot represents a competitor’s product or service.
Higher-priced, higher-benefit competitors are toward the upper right; lowerpriced, lower-benefit competitors are at the lower leƒt.
Customer-perceived benefits
100
THE McKINSEY QUARTERLY 1997 NUMBER 1
3
281
SETTING VALUE, NOT PRICE
If market shares hold constant (and if you have the right measurement of perceived benefits and perceived prices), then competitors will align in a straight diagonal line called the value equivalence line (VEL). At any desired price or benefit level, there is a clear and logical choice for customers on the VEL. So competitors aligned on the VEL say in such a market that “you get what you pay for.” The clarity of that choice almost defines a market in which shares are stable. (Note that while market shares might be stable for competitors along the VEL, their shares might not be equal. Again, more on this later.)
Exhibit 2
Share loser
E
C
A
Share gainer
VE
ad V
va
alu nt e ag e
D
L
The opposite is true for competitor E, which finds itself in a value-disadvantaged position above the VEL. Competitor E will be a share loser if the value map has been constructed properly.
Perceived price
di s ad Val va ue nt ag
e
If, however, market shares are changing, then share gainers will be positioned below the VEL in what is called a “value-advantaged” position. Competitor
A in Exhibit 2 is value-advantaged and should logically be gaining market share. If a customer is searching for a product in the benefit range of A and
B, then he or she would be more likely to choose A, since A provides the same level of benefits as B but at a lower price. Likewise, if a customer were searching for a product in the price range of A and
C, he or she would probably choose A over Value map
C, since A provides greater benefits than C
Changing market shares but at the same price. So A, positioned below the VEL that B and C reside on, oƒfers more customer value than B or C, and
B
therefore more customers prefer it.
While the marketing concepts that underpin
Customer-perceived benefits the value map are basic, advanced market research techniques (conjoint analysis, discrete choice analysis, and multi-staged conjoint analysis, for example) allow an accurate quantification of the perceived benefit dimension and its tradeoƒf against price. These advances make the eƒfective application of value maps easier than ever for marketers. That said, examples abound of costly positioning errors that could have been avoided through the use of this tool.
Illustrative case: Alpha Computer Company
The Alpha Computer Company’s experience illustrates the value map’s power, even when applied in a simple, static fashion. Alpha Computer supplied minicomputers for use primarily as servers in network applications. Alpha prided itself on its engineering skills and ability to deliver high levels of technological performance at reasonable cost. In an attempt to diagnose unexpectedly poor
THE McKINSEY QUARTERLY 1997 NUMBER 1
4
282
101
SETTING VALUE, NOT PRICE
Exhibit 3
Minicomputer value map
Alpha’s perception
Ace
market acceptance of its new line of minicomputers, Alpha created a value map that reflected its perception of the price/benefit positioning of competitors Ace Computer and Keycomp, and itself (Exhibit 3).
VE
L
Perceived price
Keycomp
Alpha believed customers chose minicomputers on the basis of two technological attributes: processor speed in MIPS (millions of instructions per second), and secondary
Alpha
access speed, that is, how quickly the computer accessed data from an external storage device such as a hard disk drive. Ace
Computer was the premium competitor: it had the highest processor speed and
Customer-perceived benefits secondary access speed, but also the highest
• Processor speed (MIPS)
• Secondary access speed price. Keycomp not only had slower processor speed and secondary access speed than Alpha but was also priced 10 to 15 percent higher. So, Alpha thought that Keycomp was value-disadvantaged and that Alpha itself was valueadvantaged.
If Alpha’s perception of the value map in Exhibit 3 were correct, then Alpha should have been gaining market share and Keycomp losing it. The opposite was occurring, however, and Alpha’s managers were baƒfled. They thought their product was superior to Keycomp’s at a lower price, and they could not understand why it was not a huge success.
Alpha’s problem was a common one. It did not understand the customerperceived attributes that really drove customer choice of minicomputers.
Alpha’s marketing department commissioned research to try to confirm its hypothesis that processor speed and secondary access speed were indeed the most important features. Sixty buyers were questioned about their criteria for selecting a network minicomputer supplier.
Much to Alpha’s surprise, processor speed and secondary access speed ranked only fourth and sixth on their list. Soƒtware and hardware compatibility, perceived reliability, and quality of vendor technical support ranked above raw processor speed. Even quality of user documents (the manual that accompanies the hardware) ranked above secondary access speed.
As it turned out, processor speed was indeed important, but most customers had a minimum processor speed requirement that all competitors easily exceeded. However, the nature of most network applications made secondary access not that important. In fact, Alpha was understood
102
THE McKINSEY QUARTERLY 1997 NUMBER 1
5
283
SETTING VALUE, NOT PRICE
by customers to be slightly better than Keycomp on processor speed and secondary access speed, but these features just did not matter that much to them.
The research also showed that Keycomp was highly rated on compatibility, reliability, vendor support, and user documents. Alpha, on the other hand, fell short on these. Its operating system soƒtware and hardware plug configuration created compatibility problems for many customers. Some remembered reliability problems with an earlier generation of Alpha’s minicomputer that tainted their perception of its new product. Alpha’s technical support Minicomputer value map was considered diƒficult to get hold of and its
Actual customer perception user documents were seen as the weakest in the industry.
Ace
Perceived price
Keycomp
Alpha
VE
L
Exhibit 4 shows how the value map was redrawn to reflect customers’ perceptions of benefits and performance rather than
Alpha’s. It showed that Keycomp performed so well on the attributes most important to customers that, despite its higher price, it was value-advantaged and therefore justifiably gaining market share. Conversely,
Alpha performed so poorly on attributes most essential to customers that, despite its low price, it was still value-disadvantaged and predictably losing share.
Customer-perceived benefits
• Compatibility
• Reliability
• Vendor support
• Processor speed (MIPS)
• Documentation
• Secondary access speed
The insights from this properly constructed value map prescribed a clear course for
Alpha. It mounted a crash program to correct the important attributes on which customers had rated it so poorly. A minor rewrite of operating system soƒtware and a simple redesign of the hardware plug configuration fixed the compatibility issue. The company then mounted an aggressive market information campaign to demonstrate the improved reliability of its latest model. Additional service representatives and toll-free access lines were put in place to enhance technical support, and user documents were redraƒted.
The results are shown in Exhibit 5. In only six months, Alpha increased customer-perceived benefits so much that it was able to increase its price by
8 percent and still gain its fair market share. The price and volume increase more than doubled Alpha’s operating profits.
The Alpha Computer case illustrates several important points about value management: THE McKINSEY QUARTERLY 1997 NUMBER 1
6
284
Exhibit 4
103
SETTING VALUE, NOT PRICE
Exhibit 5
Minicomputer value map
Repositioning Alpha
Ace
Alpha
VE
L
Perceived price
Keycomp
Customer-perceived benefits
• The key to success oƒten resides in gaining a clear understanding of the real attributes driving customer choice and their relative importance. • “Soƒter,” nontechnical attributes (perceived reliability, quality of vendor support, ease of doing business) are oƒten as important as or more important than precisely measurable technical features.
• Trusting internal perceptions of which attributes drive customer choice can be a fatal mistake; rely on customers for this critical information.
• Compatibility
• Reliability
• Vendor support
• Processor speed (MIPS)
• Documentation
• Secondary access speed
The case also shows the opportunities value maps oƒfer value-disadvantaged companies to understand their markets better. Another case, that of car maker Mazda’s experience with its Miata sports model, demonstrates the kind of opportunity that a value-advantaged company can easily forgo if it does not fully appreciate its position (see the boxed insert, “US economy sports car market, 1990”).
Distribution of customers on the value map
VE
L
Perceived price
In discussing the stability associated with a position on the VEL, and the eƒfects of competitive moves away from it, we have implicitly assumed that all positions along the line are equally attractive. This is not the case. Even for a well-defined segment, customers are not
Exhibit 6 spread evenly along the line; if they were,
Customer volume distribution every competitor on the VEL could be expected to have the same market share.
Sometimes this can be explained by hisCustomer volume torical reasons; mostly, however, it is due to the distribution of customers along the VEL
(Exhibit 6).
Customer-perceived benefits
104
History plays an important role: how long a competitor has held its position with customers oƒten explains large market share diƒferences among companies with otherwise the same value proposition. This phenomenon, also called “order of entry,” can be seen in its extreme form in deregulated utilities.
THE McKINSEY QUARTERLY 1997 NUMBER 1
7
285
US ECONOMY SPORTS CAR MARKET, 1990
I
ntroduced to the US market in 1990 at a manufacturer’s suggested retail price of
$13,800, the Mazda Miata was a retro-sports roadster that captured the imaginations of ageing baby boomer car buffs who originally fell in love with the classic British roadsters of the 1960s and 1970s made by M G and
Triumph. As much fun as its British predecessors but better built and more reliable, the Miata was an instant hit in the
United States.
Mazda Miata
Manufacturer’s suggested retail price
$ thousand
20
Mazda RX-7
18
Toyota MR-2
16
Honda Prelude
Mazda underestimated the appeal and the high perceived benefits of the simple but unique Miata. The price was disproportionately low for the perceived benefit. Mazda dealers, however, recognized this price/benefit imbalance and claimed the surplus for themselves in the form of $2,000–3,000
“ market price adjustments” that they added to the suggested retail price (and which customers gladly paid).
14
Nissan 240SX
Mazda Miata
12
0
Perceived benefit
A new competitor oƒfering similar or even slightly better value than an incumbent telephone or electricity company will not provoke the significant changes in consumer purchasing that might be expected.
A more important and probably more common explanation of market share diƒferences among competitors on the VEL is the distribution of customers along this line. Typically they are not distributed evenly, but clustered. There are several reasons for this. Sometimes consumers are not equally aware of the true nature and availability of competing products. Companies might use different channels to reach consumers, or their salesforces might not adequately communicate benefits to customers. If so, a gap can exist between customers’ perceptions of a product’s benefits and the benefits that it actually delivers.
Even in a perfect world, consumers would be unevenly distributed along the VEL because they do not necessarily view benefits and prices in a linear way. There are benefit-bracketed customers who explicitly want minimum or maximum benefit levels and find positions on either side unacceptable.
Market research shows that break-points exist for some products and services at which a small increase in the benefits oƒfered will lead to a large increase in the value a customer perceives. Some buyers of automotive components, for example, will not accept delivery reliability below a minimum level. Some computer buyers, on the other hand, do not value additional memory beyond a certain level because existing memory more than satisfies their needs.
THE McKINSEY QUARTERLY 1997 NUMBER 1
8
286
105
SETTING VALUE, NOT PRICE
A second group is price-capped customers who are unwilling to spend more than a fixed amount for a particular product or service. The price of the average home PC has held at about $2,000 for several years, even though performance has improved sharply. This could indicate that there are pricecapped customers at around this level who are unwilling to spend more even if they could get more features. Only customers who fall into neither category, benefit-bracketed or price-capped, are actually willing to consider the full range of tradeoƒfs along the VEL.
Understanding volume distribution along the VEL is therefore crucial to making an intelligent decision about product position. In many cases, however, it is poorly understood, leading to wrong decisions. Typical mistakes are:
• Positioning an apparently competitive product at a low-volume part of the VEL and not getting the expected volume gains. A maker of metalcoating machinery positioned a new product technically half way between two competing products, hoping to pull in customers not entirely satisfied with these. What it had not realized was that there was no significant volume between the two extremes, because each answered a specific speed requirement of downstream customers. Failing to understand that there was no demand for a medium-speed machine, even one that was competitive on technical specification and price, forced the manufacturer to take a multimillion-dollar writeoƒf.
• Positioning a product too high or too low on the VEL, thereby inadvertently excluding a large portion of price-capped or benefit-bracketed customers.
The drastic fall in demand for one company’s supercomputers is an example of this. Even though the company’s ever more powerful machines remained on the VEL, there was no longer a customer imperative for all that processing power to be concentrated in one machine, as more broadly distributed processing had become preferred by most users.
Dynamic value management
Alpha Computer and Mazda Miata illustrate the pitfalls of failing to understand the “static” value positioning of a product or service. But getting a product to the right position on a static value map is only part of managing value eƒfectively. Unfortunately, neither competitors’ positions on a value map nor customers’ perception of products and suppliers are frozen in time.
Value maps are not static but dynamic, constantly changing in important and oƒten predictable ways.
Any change in product positioning by one competitor, be it cutting price or improving features, will lead others to move, either to preempt shiƒts in market share or to react to them. We apply the term “dynamic value manage-
106
THE McKINSEY QUARTERLY 1997 NUMBER 1
9
287
SETTING VALUE, NOT PRICE
ment” to the discipline of managing price/benefit positioning not just in a static fashion, but with explicit and thoughtful consideration of likely changes in competitive value positions and customer value perception.
Companies that master this discipline can reap huge rewards and avoid equally huge pitfalls.
Illustrative case: MTE
Perceived price
MTE is the manufacturer of high-quality medical testing equipment mentioned at the beginning of this article. Its primary product was a blood diagnostic testing machine used in highExhibit 7 volume hospital laboratory applications. Dynamic value management
MTE was the recognized premium supplier
Leader enhances benefits but does not raise
(with the highest price and benefits) in a price commensurately… stable market that included three other leading competitors (Jackson, PZJTech, and
Labco) positioned squarely on the VEL
MTE
(Exhibit 7).
Jackson
VE
L
As is oƒten the case, MTE, as the premium
PZJTech
supplier, was the real innovator in this market. The improved version of its blood
Labco
diagnostic testing machine was more accurate and had faster testing cycle times.
But MTE was in a dilemma over how to price its terrific new model. Research showed
Customer-perceived benefits the added benefits would justify a 10 percent price increase and still keep the model on the
VEL – that is, MTE would hold its market share. But, equally, it could keep the price the same and position the new model in a highly value-advantaged position in the hope of gaining significant market share.
MTE decided on a compromise, raising its price by 5 percent, a meaningful increase that still kept it in a value-advantaged position (the dotted circle in
Exhibit 7). The response was instant and positive. Customers recognized the
5 percent increase was a small premium to pay for enhanced accuracy and cycle times. The machine sold well and immediately increased MTE’s share of the market.
This success, of course, was at the expense of Jackson, PZJTech, and Labco, none of which had the expertise or resources to introduce products to rival
MTE’s new model. Faced with falling sales, they took the only measure they could to defend their market shares – they lowered their prices by at least
5 percent (Exhibit 8). The market shares of all four companies quickly returned to their previous levels, but at the lower prices. As Exhibit 8 shows, the VEL had simply shiƒted downward and MTE’s value-advantaged position was
THE McKINSEY QUARTERLY 1997 NUMBER 1
10
288
107
SETTING VALUE, NOT PRICE
Exhibit 8
Dynamic value management
O ld VE
L
…resulting in a downward price shift across market
VE
L
Jackson
N ew Perceived price
MTE
essentially nullified. The lowered
VEL was good for customers because they got more for their money, but the suppliers got less for their products. It represented a wholesale transfer of market surplus from suppliers to customers.
Could MTE have managed the value dynamics of this situation better?
Possibly. If it had raised the price of
Labco
its new model by 10 percent and positioned it on the existing VEL, it would have held its traditional share but at a 5percenthigherprice.Jackson,
Customer-perceived benefits
PZJTech, and Labco, experiencing no loss of market share, would probably not have reacted at all. Industry prices would have been maintained, and
MTE’s profit would have risen significantly.
PZJTech
Changing your position in a dynamic world
Marketing managers have two basic options for improving their products’ position, regardless of whether they are in a proactive or reactive situation.
They can reposition their product along the VEL, or move oƒf it. These diƒferent moves engender very diƒferent outcomes – diƒferent competitor and customer reactions and diƒferent prices, volumes, profits, and risks.
Repositioning along the VEL
Repositioning a product along the VEL, usually a less aggressive move, requires a company to understand where customer clusters are on it, and how other competitors are positioned in relation to them. The decision of whether and how far to move should include the following steps:
Understanding and weighing the risks and opportunities. Repositioning a product is likely to lose some customers who preferred the old positioning.
Equally, it will gain customers who prefer the new positioning. Failure to understand this tradeoƒf could lead a company to surrender a good customer franchise in exchange for a reduced, and probably more competitive, new franchise. Being smart about choosing the right attributes to vary. Customers do not consider all product attributes to be equally important; there is therefore more “bang for the buck” in changing some attributes rather than others.
The knack is to select the features that will attract new customers without
108
THE McKINSEY QUARTERLY 1997 NUMBER 1
11
289
SETTING VALUE, NOT PRICE
losing old ones, that have the greatest impact on customers, and that the company can provide cost-eƒfectively.
Knowing what price change is appropriate for a given attribute change.
If the aim is to stay on the VEL, any change in benefits must be accompanied by a price change. Not increasing the price enough will force competitors to match the new positioning, leading to an unwanted industry price decline
(as with MTE); raising the price too high will lead to a volume loss. Market research tools such as conjoint analysis can determine the magnitude of change required.
Choosing those changes least likely to provoke undesirable competitive reactions. If the repositioning is successful, or looks as if it will be, competitors will react. The likeliest, and least desirable, reaction is a price cut, which oƒten leads to price cuts across the industry and lower profits for all. One manufacturer of medical supplies always reacted to competitors’ price cuts by improving benefits. Every time a competitor dropped its price, the supplier countered with an improved version of its product at the same price, but on the new VEL. In this way it gained a distinctive market position, offering increasingly superior benefits over competitors that chose to move only along the price dimension.
Choosing the new position along the VEL. There are two options: either to move to a new position within the extremes defined by current competitors, or to move to a new position beyond the current extremes. There are diƒferences in risk and potential competitive moves between the two:
• The success of a new positioning within current competitive extremes depends on locating the right customer concentration and standing out from competitors. As this approach seldom expands a market, competitors will probably react to their declining sales.
• Moving to a new position along the VEL outside the existing extremes can expand a market. While the upside opportunities can be greater (and the threat of retaliation lower), success depends on a thorough understanding of the size and needs of the latent demand that the new product or service is designed to meet.
Moving oƒf the VEL
A move oƒf the VEL into value-advantaged territory might seem attractive on the surface. As the experiences of many companies show, however, such a move requires an even better understanding of the dynamics, risks, and opportunities than do moves along the VEL.
What is diƒferent about moves oƒf the VEL? A repositioning along the VEL is likely to threaten only one or two neighboring competitors currently on the
THE McKINSEY QUARTERLY 1997 NUMBER 1
12
290
109
SETTING VALUE, NOT PRICE
Perceived price
line. Moving below the VEL oƒten threatens all competitors, because such moves usually define new and lowered VELs that force them to reconsider their own positions. Only rarely does the
Exhibit 9
VEL move upward; to do so would require
Repositioning off the value equivalence line customers to accept the actual value reduction and most suppliers to move in the
Expanding the horizon same direction.
Addressable customer horizon
A
VE
L
A
When a product is repositioned below the
VEL, its “horizon” of potential customers grows (Exhibit 9). Take, for example, an electric drill whose power was increased but which was sold for the same price. The new product appeals not only to customers who initially bought it, but also to those who had previously paid more for a drill with the higher power rating.
Just moving oƒf the VEL to expand the horizon of customers does not guarantee success, however. Market research must first establish that the expanded horizon does indeed include new concentrations of customers, not just empty space.
Customer-perceived benefits
Likely competitive reactions to moves oƒf the VEL
In today’s highly competitive markets, rivals seldom passively accept volume or market share losses. They usually react by trying to improve their products by selectively adjusting attributes, or by dropping price. How they will react is a function of a number of parameters, including:
The type of change that set the whole process in motion. The typical reaction to a competitor’s move is to try to counter along a similar axis. If the salesforce reports massive price cuts by a competitor, they will want to reciprocate. If a competitor introduces a new service, the salesforce will want to oƒfer something similar. A first mover’s repositioning along the benefits axis tends to damage profits less than price reductions would. It is also easier to retract benefits that are rejected by the market or are uneconomic to provide, than to try to raise prices aƒter a round of reductions.
Competitors’ strategic mindset. The degree of volume and profit pressure a competitor is under and its understanding of the economics of price changes (for example, how price and volume trade oƒf against profit) will drive the type of reaction it makes.
Even in commodity-like industries, there are examples of manufacturers successfully improving their products and services rather than cutting prices.
110
THE McKINSEY QUARTERLY 1997 NUMBER 1
13
291
SETTING VALUE, NOT PRICE
In a US specialty chemical segment, for example, the two leading companies have about 40 percent of the market. They and their customers recognize that there are no real technical diƒferences between the two suppliers’ products. So when one competitor increases its support services, the other improves its services too. While the industry is competitive, and the level of service high and rising, prices have also risen and profits have remained strong. In the past five years, neither leader has reacted to a competitor by reducing its price – a move that would surely have made the industry less profitable. Eƒfects on demand and volume distribution
Competitors’ behavior can actually shiƒt the distribution of demand along the VEL (Exhibit 10). As the line is shiƒted downward through improved combinations of price and benefit, it is not automatic that the “old” pattern of customer distribution follows suit. Some customers might be benefitbracketed, others might use the changes to rethink their own price/benefit tradeoƒfs, and, finally, new oƒfers could stimulate latent demand.
Repositioning off the value
Exhibit 10
equivalence line
ld
N
ew
VE
L
3
O
2
VE
L
“Old” and “new” customer volume distribution
Perceived price
If the distribution of demand changes, a shiƒt oƒf the VEL will not always bring the desired volume increase. The established manufacturers in one consumer durable industry assumed most customers were price-capped, and therefore had not oƒfered increased benefits. But when a new competitor introduced a new product at a significantly higher price, 30 percent of the volume shiƒted to that new product. Some consumers had been looking for more benefits aƒter all.
1
Customer-perceived benefits
A move oƒf the VEL has to be large enough for customers to notice and attractive enough to make them want to try the repositioned product. Marginal moves oƒten backfire. If consumers do not perceive enough diƒference to make them switch supplier, but competitors, which follow such moves closely, decide to copy it, the VEL can quickly drop without aƒfecting market shares, but lowering price and profit.
In the case of a company that installed heating equipment, the information that its key competitor had cut the cost of installation labor by 5 percent led it to cut its own price too. Unfortunately, this company did not adequately consider the basis on which architects and contractors compare bids – that is, the total installed costs. The selective 5 percent drop in labor reduced the
THE McKINSEY QUARTERLY 1997 NUMBER 1
14
292
111
SEGMENTATION THEORY
I
n classic marketing and segmentation theory, a segment is defined as a group of customers with identical needs and buying behavior. In theory, all customers constituting a precisely defined segment would be equally receptive to all products positioned anywhere along the
VEL. Therefore, all products positioned on the
VEL should have the same market share.
In practice, this is clearly not the case.
There are two ways to resolve the conflict:
• Define a segment (and the products positioned in it) so that it contains all the realistic and feasible alternatives customers consider for a given purchase, and accept that there will be some differences in buying decisions. We will take the second approach, as it makes the concepts in this article easier to apply in the messy “real” world, without compromising the quality of the answer.
• Define each segment so narrowly that it contains only one customer.
total installation cost by less than 1 percent – too slight a diƒference for the market to notice.
Moving oƒf the VEL therefore requires two decisions about the direction and the distance:
• Direction. What are the customer volume elasticities of moves along the price axis and the benefit axis (by attributes)? Do I want to increase my benefits, lower my price, or both?
• Distance. How far do I have to move from the VEL to expand my horizon of customers suƒficiently? How far do I have to move to diƒferentiate myself from competitors in the eyes of a group of potential customers? How strong will competitors’ reactions be? How many additional benefits can I aƒford to deliver and what price cut am I willing or able to absorb?
Moving below the VEL is always a risky strategy that can, if executed well, reap some benefits. In many cases, however, too little thought is given to what customers actually want, how competitors will react, and how demand might change as a result of competitors’ moves. This negligence can lead to profit declines where once there were high hopes.
Using dynamic value management to respond to external changes Dynamic value management can also be a powerful tool to help prescribe reactions to changes in competitive position or customer needs. A competitor’s actions can set in motion the same set of dynamics. Dynamic value management is as useful in determining reactions to such moves as it is in initiating them.
Competitors’ moves
Being on the receiving end of a competitive move demands an approach similar to the proactive stance above. It also requires a cool head. If the
112
THE McKINSEY QUARTERLY 1997 NUMBER 1
15
293
SETTING VALUE, NOT PRICE
salesforce is sending panicky messages about competitive price cuts, pressure is created to act quickly. In most cases, the easiest lever to pull in the short term is price. And in all too many cases, this would be a mistake. A series of thoughtful decisions using the dynamic value-management approach can help formulate a more eƒfective and less costly response. A set of questions should be answered:
• Do customers perceive the competitor’s move as a move oƒf the VEL?
To find out, ask the customer. Too oƒten this question is answered hastily and wrongly on the basis of hearsay from the field. If the move is not perceived to be a wholesale jump to a new VEL, there may be no need to react.
• If the competitor has moved oƒf the VEL, has its “horizon” expanded suƒficiently to draw in new customers? If market research shows it has not, again there is no need to react.
• If new customers are buying the competitor’s oƒfering, are they our customers or somebody else’s? The answer to this question determines not the need for a reaction, but the speed and extent of it. If the primary threat is to somebody else’s customers, let them react. All competitors will be likely to react eventually, but timing is important. A gradual cascade of reactions not only will prevent panicky overreactions, but can also create opportunities to observe informative customer buying behavior.
• If a reaction is needed, how strong should it be? Should it be a surgical strike on one product, channel, or market, or across the board? Should it entail price changes, benefit changes, or a combination of both?
Cyclical markets
In cyclical industries, the value map can change not only because of competitors’ moves, but also because of changes in customer needs over the course of the cycle. The following case illustrates the enhanced challenge of dynamic value management in highly cyclical businesses.
Illustrative case: Pace Paper Company
The Pace Paper Company produced high-grade paper for business forms, brochures, and corporate annual reports. Pace and its two main competitors,
Marco Paper and Valentine Paper, sold directly to large regional and national printing companies. Demand for this high-grade paper tended to vary wildly with the overall economic cycle.
Pace produced paper of unsurpassed quality and consistency and provided equally consistent delivery service. But it found itself gaining market share in down markets when there was excess supply and losing it sharply in times of tight supply. Given its consistency and quality throughout the
THE McKINSEY QUARTERLY 1997 NUMBER 1
16
294
113
SETTING VALUE, NOT PRICE
economic cycle, Pace could not understand the drastic market share shiƒts it regularly experienced.
The problem was that while Pace stayed the same, customers changed through the cycle. Exhibit 11 shows the value map for this market at diƒferent stages of the cycle. At the bottom of the cycle (when supplies were plentiful),
Exhibit 11
Value position changes through a cycle
Industry supply /demand balance
Excess supply
Tight supply
Stable market shares
Benefit attributes ranked
1. Paper quality/consistency
2. Order lead time
3. Order fill rate
Pace
Marco
Valentine
Pace
Marco
Benefit attributes ranked
1. Order lead time
2. Order fill rate
3. Paper quality/consistency
Valentine
Time
customers had no problem obtaining enough paper. They therefore demanded high and consistent quality so that printing jobs would run eƒficiently through their plants with minimum rejects. Order lead times (the number of days between an order being placed and the paper being delivered) and order fill rates (the percentage of the total order carried on the first shipment) did not much matter, since printers usually had ample safety stocks of paper in their own warehouses. Despite being slightly higher priced than Marco and Valentine, Pace’s paper quality and consistency were so superior that it was value-advantaged in times of excess supply and gained market share – as shown at the lower leƒt of the value map in Exhibit 11.
As supplies tightened, however, printers oƒten found their stocks depleted.
They became increasingly concerned about running out and having to shut down their printing plants temporarily. They therefore relaxed quality and consistency requirements in favor of delivery performance. As the value map at the upper right of Exhibit 11 shows, paper quality and consistency slipped to third place behind order lead time and order fill rate. Valentine Paper could not match Pace’s quality and consistency, but its order lead times and fill rates were better than Pace’s. The result was that in times of tight supply,
Valentine would shiƒt to a position of value advantage (and thus gain market
114
THE McKINSEY QUARTERLY 1997 NUMBER 1
17
295
SETTING VALUE, NOT PRICE
share), while Pace would slip to a value-disadvantaged position (and, of course, lose share).
Armed with the insights provided by the value maps in Exhibit 11, Pace’s managers embarked on a project to determine how they might improve their order lead times and fill rates in times of tight supply. They discovered that if they relaxed their product consistency slightly (in a way that was almost imperceptible to customers), they could increase plant throughput enough to cut order lead times, increase order fill rates, and return to value equivalence in tight markets. When supplies loosened, Pace reverted to its original level of paper consistency to reinforce its traditional valueadvantaged position. This fine-tuning over the course of market cycles enabled Pace to maintain its share in tight markets without cutting prices or jeopardizing future positioning in down markets.
Looking ahead
With product life cycles shrinking (measured in months rather than years in the computer industry, for instance), customers becoming more sophisticated and demanding, and tougher local and even global competitors emerging in most markets, value maps are shiƒting at faster rates than ever. Fortunately, advances in market research techniques make the execution of eƒfective dynamic value management easier than ever.
The discipline of dynamic value management not only promotes sustainably improved market performance and profitability, but also yields a number of attractive side benefits, including:
• More genuine closeness to customers, thanks to a richer, more externally driven understanding of the benefit attributes that really matter to customers
• An enhanced understanding of competitors: their strengths in the eyes of customers, their strategies, and their likely reactions to price and benefit moves by your company
• More integrated product/market strategy formulation, where the linkages between price, benefit delivery to customers, competitor capabilities, and changing customer preferences are explicit.
The payoƒf for getting dynamic value management right has probably never been higher; the consequences of getting it wrong, never more devastating.
For a growing number of companies, dynamic value management is providing a compass for navigating the increasingly unstable seas of change and uncertainty that challenge most marketers today.
THE McKINSEY QUARTERLY 1997 NUMBER 1
18
296
115
IMB 425
SUHRUTA KULKARNI, KRIPA MAKHIJA AND U DINESH KUMAR
APOLLO HOSPITALS: DIFFERENTIATION THROUGH HOSPITALITY
The ‘‘wow’’ factor in service relies on constant innovation and demands continuous and sensitive focus on all issues that may affect the patient’s stay in a hospital. Every touch point of the hospital needs to be ‘‘alive’’ and the client must be able to feel the warmth offered. The culture of service is imperative in today’s scenario, where the differentiators could just be the manner in which services are offered. All the major players could replicate infrastructure within a short span of time, but not the service culture.
Dr. Umapathy Panyala, Chief Executive Officer, Apollo Hospitals, Bangalore (March 2013)
Dr. Panyala, CEO, Apollo Hospitals, Bangalore believed that in the future, the hospitality aspect of hospitals—the service provided to patients—would differentiate Apollo Hospitals from a large number of equally competent competitors in the growing Indian healthcare industry. He had set up a quality department at the Apollo Hospital in
Bangalore, headed by Dr. Ananth Rao. Apart from being an expert on Metabolic Diseases and Biochemistry, Dr.
Rao was also a Lean Six-Sigma black belt from the Indian Statistical Institute, Chennai.
You can’t manage what you don’t measure—although this may sound clichéd; I am still a firm believer of this philosophy and want to apply this, especially in the hospitality part of hospitals.
Clinical benchmarking is a compulsory requirement and is taken care of; however, patients have so many other touch points in their stay at hospitals—the hospitality part. Some of the world-class hotels (such as the Ritz–Carlton) have performed benchmarking to standardise their hospitality offerings; at the same time, its employees are allowed to use their judgment to provide whatever delights the customer in every visit. 1 We want to internalise this in our culture as well.
– Dr. Ananth Rao, Head–Quality Department, Apollo Hospital, Bangalore (March 2013)
Dr. Rao also believed that the hospitality aspect would differentiate Apollo Hospitals from its competitors. Patient cure and care played very important roles in hospitals. Many patients were generally anxious when in a hospital and the sense of disservice increased their anxiety level very easily. Integrating healthcare and hospitality was essential for creating patient-focused care. Hospitality aspects included a smooth admission procedure, friendly medical and non-medical staff, comfortable furniture, varied choices on the food menu, attractive surroundings, recreational facilities, privacy, clear signposting, adequate provisions for visitors, and so on. 2 Important aspects of hospitality were managed by the nursing staff and other non-medical staff, which meant inherent variability of service owingto human interventions.
Dr. Panyala and Dr. Rao wanted to measure the hospitality aspects at Apollo Hospitals and improve hospitality to create a world-class hospital. Dr. Rao and his team collected feedback every day from the patients and received a number of complaints, ranging from not having a TV remote to long response time on the part of nursing staff in attending to requests from patients. For Apollo Hospitals, it was important that the patients’ experience in the hospital was not compromised, since it could have a significant financial impact. Managing the hospitality elements of the hospital was as important as managing the clinical aspects. Apollo Hospitals had a stringent process in place to take care of clinical aspects. Dr. Rao wanted to improve the hospitality at Apollo Hospitals by reducing the
1
Hall, J. M. and Johnson, M. E., When should a process be art, not science, Harvard Business Review, 2009, 1–9.
Hepple, J.,Kipps, M. and Thomson, J., The concept of hospitality and an evaluation of its applicability to the experience of hospital patients,
International Journal of Hospitality Management, 1990, 9(4),305–318.
2
Suhruta Kulkarni, Kripa Makhija and U Dinesh Kumar, Professor of Quantitative Methods and Information Systems, prepared this case for classroom discussion. V Sandeep assisted in data collection and analysis. This case is not intended to serve as an endorsement or source of primary data, or to show effective or inefficient handling of decision or business processes.
Copyright © 2013 by the Indian Institute of Management Bangalore. No part of the publication may be reproduced or transmitted in any form or by any means – electronic, mechanical, photocopying, recording, or otherwise (including internet) – without the permission of Indian Institute of
Management Bangalore.
19
Page 2 of 20
Apollo Hospitals: Differentiation through Hospitality
number of complaints from patients; he also wanted to achieve significant improvement in sigma levels measured through the Six Sigma performance scale. According to Dr. Ananth Rao:
The immediate challenge is to understand the patients’ sentiment towards the hospitality provided and to design a process improvement plan that is affordable. Apollo takes feedback from patients every day and the quality department staff interviews many patients every week to understand their needs.
Dr. Rao was aware that improving hospitality at Apollo Hospitals was going to be a continuous exercise in improvement; collecting feedback was one way of approaching the process of continuous improvement. He treated every complaint as a “defect” and planned to use lean Six Sigma concepts to eliminate defects. Implementing Six
Sigma in all departments was likely to be a challenge since departments such as housekeeping faced high attrition rates. His immediate objective was to introduce a system where future complaints related to hospitality could be reduced.Also, how much importance should be given to hospitality by Apollo Hospital was one of the dilemmas faced by Dr Rao and he wanted to set a realistic target for Sigma level in hospitality at Apollo.
APOLLO HOSPITALS: THE TRENDSETTER
Dr. Prathap C. Reddy, founder of Apollo Hospital Enterprises Ltd. (AHEL) had accomplished a successful medical career in the United States. He returned to India in 1972 to contribute to the healthcare system in India. Health infrastructure in India was very poorly developed in the 1970s. In 1971, there were 3,862 hospitals and 12,180 dispensaries with a total of 348,6553 beds for a population of 548,159,6524—a ratio of 6.36 beds per 10,000 people as against the ratio of 9 beds per 10,000 people in 2011. 5 India’s first National Health Policy was declared in 1983,6 almost 36 years after independence, which was an indication of the neglect faced by the health sector in the country since independence.
Dr. Reddy had set up a good medical practice in India and used to send patients outside the country for specific treatments. However, in 1979, a young patient died as he could not arrange the money for treatment in the United
States. Dr. Reddy then decided to provide the best of medical treatment from the West to patients in India with an emotional touch, calling it “High Tech with High Touch.” Apollo was a doctor-promoted enterprise—10,000 Indian doctors, 4,700 U.S.-based doctors, and 60 doctors from the United Kingdom invested approximately USD 5,000 to start the venture. Dr. Reddy selected the best of the talent available to ensure the best possible service and care. He also ensured that a clear distinction was maintained between business management and clinical management. 7Apollo pioneered world-class healthcare in India, which was later emulated by several other hospitals.
Apollo focused on technological excellence and garnered many firsts to its credit in the country. Apollo was the first not only in India but also in South Asia to launch Oncological Robotic Surgery, G4 Cyberknife Robotic
Radiosurgery System, 320-slice computed tomography scanner, 64-slice positron emission tomography-computed scan system, full-field digital mammography with tomosynthesis, and many such technologies. 8 According to Dr.
Rao, Apollo intended to carry forward technological excellence in hospitality to provide patients with the best cure and care services.
Dr Preetha Reddy, Managing Director, Apollo Hospitals Enterprises Limited has been the pioneer and chief architect of the tender loving care –TLC ‘‘mantra’’, a pillar of the Apollo way, which is affectionately applied to every patient at Apollo Hospitals. “The patients and staff comprehend this language better,” she points out. The concept of TLC integrates service delivery with clinical outcomes resulting in exceptional patient experiences9,10.
3
Background Papers: Financing and Delivery of Healthcare Services in India, National Commission on Macroeconomics and Health, Ministry of
Health and Family Welfare Government of India, 2005, p. 47.
4
Source: http://cyberjournalist.org.in/census/cenpop.html, accessed on March11, 2013.
5
Source: http://www.globalhealthfacts.org/data/topic/map.aspx?ind=78, accessed on March11, 2013.
6
Health Research Policy, Indian Council of Medical Research, New Delhi, (October 2007).
7
Mitra, M., The Apollo Mission, Corporate Dossier with The Economic Times, June 1, 2012.
8
Apollo Investor Presentation, www.apollohospitals.com, accessed in January 2013.
9
N Amarnath, and D Ghosh, The Voyage to Excellence: The Ascent of 21 Women Leaders of India Inc., Pustak Mahal, pp. 80-95.
10
http://www.apollohospitals.com/apollo_pdf/dr_preetha_reddy_managing_director.pdf
20
Page 3 of 20
Apollo Hospitals: Differentiation through Hospitality
CLINICAL BENCHMARKING
Apollo Hospitals had been using a clinical score card called ACE@25 (Apollo Clinical Excellence), which measured and monitored clinical excellence among the group’s hospitals. ACE@25 measured 25 clinical parameters
(Exhibit 1) every month, which were benchmarked against global standards. ACE@25 was launched on
September18, 2008 and used across 32 hospitals of the group. Clinical benchmarks were published by various institutions and bodies such as Cleveland Clinic, Mayo Clinic, and National Healthcare Safety Network (NHSN), among others. Hospitals were grouped according to their bed strengths, locations, services offered, and so on. Group
A hospitals had to report 25 parameters—23 were common parameters, while two were location-specific. Group B and Group C hospitals had to report 15 and 10 parameters, respectively, out of which two were location-specific.
ACE@25 was an internally developed clinical scorecard, created by drawing upon the wealth of expertise available within Apollo. An audit committee at the corporate level was set up to validate the data, methodology, and definitions followed at each location. According to Sangita Reddy, Executive Director, Apollo Hospitals Group:
We needed a yardstick like ACE@25 that would pit us against international institutes like
Cleveland Clinic, Mayo Clinic, and others, and position us on the global healthcare firmament for excellence in clinical quality. This also enables us to assess where we stand and where we need to
11
be, while pursuing excellence in clinical quality.
Apart from this internal benchmarking exercise, seven of Apollo’s hospitals were accredited by the Joint
Commission International (JCI); and it was the largest group in South Asia to be accredited by the JCI. The JCI was a U.S.-based accreditation body dedicated to improving healthcare quality and safety around the world and recognized as the gold standard for hospitals. Apollo was also accorded the Superbrand status by the Indian
Consumer Superbrands Council, which recognised that the best practices were used in the brand. Apollo was the
12
only hospital that was accorded the Superbrand status in India. There were other accreditations that several Apollo hospitals had achieved (Exhibit 2).
According to Dr. Panyala,
Living the brand should be our focus in every initiative or activity we perform. Apollo Hospitals has been one of the consistent names among the Superbrands. The perceived value of a brand like
Apollo Hospitals is set very high in the backdrop of the decades of service and excellence it has offered. Clients need to see and experience that value, and the gap between perceived value and obtained value must be zero at best or at a bare minimum.
PATIENTS’ FEEDBACK AND REAL-TIME ACTION
On average, a patient spends 80% of the time in hospital for the care part rather than the cure, and we need to focus on care to ensure speedy recovery and maximum satisfaction. Hospitality is critical in healthcare as the patient and his/her attendants are already distraught and highly anxious. Hospitality is driven mainly by human interventions—in nursing, housekeeping, as well as food and beverages. It is very difficult to ensure consistency of quality and hence, we want to benchmark these to ensure we provide the best quality of hospitality all the time.
–Dr. Ananth Rao, Head–Quality Department, Apollo Hospital, Bangalore (December 2012)
Dr. Rao believed that although clinical services formed the core of Apollo’s services and brand image, hospitality would support the brand, and in the long run, both would merge to form the Apollo brand (as shown in Exhibit 3).
All services that did not require core clinical expertise were classified as hospitality services, including services such as billing, dietician service, food & beverages, facility, housekeeping, nursing, facility, and overall operations. Each service was executed through a variety of processes. All the processes included in each service were identified and defined with regard to the procedure, timelines, required output, and so on. All the processes were mapped and the quality measures defined; these would be used as Sigma metrics.
11
12
Express Healthcare, (2010), http://healthcare.financialexpress.com/201009/strategy01.shtml
Source: http://kolkata.apollohospitals.com/newsroom/271-apollo-hospitals-only-healthcare-super-brand-in-india.html
21
Page 4 of 20
Apollo Hospitals: Differentiation through Hospitality
The Quality Department, established under the leadership of Dr. Rao, comprised two dedicated staff—Soumi Dutta and Nisha Maria—who looked after a variety of quality-related issues. Soumi and Nisha collected feedback from the patients between March 2011 and December 2012using the form presented in Exhibit 4. Patients were asked to rate each department on a scale of 1 to 10. Additionally, open-ended feedback such as patients’ comments, opinions, or suggestions was also collected.
A schedule was developed for collecting feedback, which ensured that Soumi and Nisha collected feedback from a cross-section of patients; this also ensured that no biases crept into the feedback. The feedback collection methodology is shown in Exhibit 5. Soumi and Nisha were trained to collect frank, free-flowing feedback from the patients. If they received complaints while collecting feedback, they would immediately inform the department concerned and get the errors rectified, whenever possible; or ensure that the complaints were addressed to the patient’s satisfaction in real time. One of them recollected the following anecdote:
A patient had complained that the door was not getting locked properly. I got in touch with the facility personnel and they worked on the door and the lock and fixed the problem—all in a matter of 25 minutes from the moment it was brought to my attention. The patient was satisfied with the immediate solution. However, we did not stop there. We teamed up with the facility team and checked every door of the hospital and repaired them if required. We wanted to ensure that such complaints were not repeated.
The feedback collection process served multiple objectives such as collecting open-ended feedback from patients, resolving the issues in real time, and further auditing the actions of the service departments. The real-time escalation flowchart is shown in Exhibit 6.
The feedback was saved on an MS Excel spread sheet, and stored on a monthly basis for easy retrieval. The data was then analyzed using various parameters and trends were plotted for each service. At Apollo, each service was related to a department; hence, it was easier to deal with the complaints and determine monthly improvements.
FEEDBACK ANALYSIS
From March 2011 to December 2012, 1,434 complaints were received from among the 1,38,600 in-patients treated during that period (approximately 1.03%). A Pareto chart was plotted for these complaints (as shown in Exhibit 7).
The housekeeping department received the maximum number of complaints, while the dietary service had the least number of complaints. The department-wise spread of complaints is shown in Exhibit 7. Some of the complaints were genuine concerns while some were related to minor discomfort. A few of the complaints were very specific, while some were generic. All of these were analyzed, which would enable the hospital to work towards reducing the overall number of complaints. Some of the sample complaints from each department are provided in Exhibit 8.
According to Dr. Rao,
Every complaint is an opportunity to improve. We keep looking for the smallest of the complaints, which will help us in improving our quality by several levels. Sometimes it is difficult to interpret the complaints and it is even more complex to develop strategies that will enable a better patient experience. The complete data set was analyzed to determine the word frequency count in the complaints section. The significant words with their frequencies are shown in Exhibit 9. This analysis was used to focus on specific tasks to ensure reduction in the number of complaints. For example, the most significant word was “time” and it was associated with delays in response time for the various services. The twenty-fourth most frequent word was “late,” which is again related to response time. Thus, the word frequency technique helped in focusing on problem areas.
Based on the results of the analysis, benchmarks were set in consultation with the respective department for the response time of each service as shown in Exhibit 10.
Apart from this quantitative analysis, another approach was used to analyze the feedback and obtain deeper insights for quality improvements. Dr. Rao used the term “defect-defective” from the Six Sigma methodology—one
“defective” product/service could be caused by one or several “defects”. According to Dr. Rao,
22
Page 5 of 20
Apollo Hospitals: Differentiation through Hospitality
Any complaint from a patient is considered as a ‘‘defective’’. For example, consider the complaint: ‘‘Food is not served on time’’. This complaint may arise due to several reasons such as food not being prepared in time, food not being delivered on time, patient changing his/her order, etc. It is essential to identify these defects in order to eliminate the defective.
On receiving a complaint from the patient, which was termed as “defective,” defects that led to the defective
(complaint) were identified. Root-cause analysis was performed on all the processes of the identified defects. The processes were re-engineered to eliminate all the defects and a pilot study was conducted using the “Define Measure
Analyse Improve Control” (DMAIC) cycle. Once the process was found acceptable, it was then deployed across locations. This was followed by routine and surprise audits to ensure that the process was being followed as defined to ensure customer satisfaction. The flowchart is shown in Exhibit 11a and b. All feedback related to medical services was escalated to the Medical Director’s office.
In addition to this, the Quality Department at Apollo Bangalore developed a methodology called the Daily Point
Average© or DPA©. The ratings provided by patients for different departments were used to calculate the DPA©. The departments had to improve these ratings over a period of time. The DPA© effectively captured the “mind of the customers” since the feedback was collected during the patients’ stay and not at the time of discharge.
BENCHMARKING OF HOSPITALITY
Hospitality required high human involvement and was very specific not only to local cultures but also to each individual. Since a patient had to stay in a hospital to get cured, hospitality automatically came into the picture.
Hospitality in various hospitals was very different owing to the surroundings and differences in customer (patient) requirements. Patients did not walk into a hospital out of volition—they came in only because there was some problem. Under such conditions, the patient would be very agitated and any small thing that was out of place would become a big issue. Any kind of delay would be extremely intolerable and all the services had to be perfect all the time. Even in the hotel industry, hospitality was not standardized and benchmarks were not available. The Ritz–Carlton hotels, which are considered the gold standard in the hospitality industry, had used Six Sigma and benchmarking for their hospitality business.13 Although benchmarks for clinical services were well-established, those for hospitality in hospitals were yet to be established.
Apollo Bangalore developed benchmarks for several common complaints with three levels of services (as shown in
Exhibit 10) by adopting the Kano model, which was developed by Noriaki Kano (Exhibit 12). This model was used across service industries and it helped in understanding customer expectations from any product or service. The threshold or the basic quality was the minimum requirement of the customer, which would be taken for granted even if it were present; however, if it were not there, the customer would complain about its absence. Normal or performance quality was something that the customer would expect because these were stated either by the product/service provider or by the customer as a requirement. This quality was observed by the customer and its absence would cause discomfort and disappointment. Exciting quality of the service or product was something that would not disappoint the customer; the presence of this would delight the customer, since the customer did not expect this quality. With time, the exciting quality would become performance quality and the performance quality would become a basic quality. Hence, the manufacturer or the service provider should always strive to provide new exciting qualities.
Accordingly, several metrics were defined for benchmarking. For example, patients were informed that routine hospital-provided meals would be served within 10 to 20 minutes of every mealtime. This became a performance attribute. The threshold requirement of the patient would be that meals should be served within 20 minutes after placing the order. If the meal was served within 10 minutes, the patient would be delighted. However, if this customer (patient) were to come to the hospital again, she/he would expect the meal to be delivered within 10 minutes; this then would become a performance quality for her/him.
13
Source: http://www.qfdi.org/newsletters/six_sigma_qfd_hotel_application.html
23
Page 6 of 20
Apollo Hospitals: Differentiation through Hospitality
FINANCIAL IMPLICATIONS
The Quality Department required funding for data collection, analysis, and other related activities. These expenses affected the bottom line of the hospital. However, from March 2011 to December 2012, the number of re-visits increased, which implied that customer loyalty had increased. There was a 15% increase in the number of inpatients. Earlier, high discounts had been offered to dissatisfied patients owing to errors in service or poor quality of service. Gradually, there was a reduction in the discounts provided, which was a direct result of better satisfied patients. Additionally, owing to better processes, cost of quality (in terms of re-work and consumable wastage) had reduced, which helped in improving the bottom line. Further, owing to better service and higher levels of satisfaction, the patients acted as brand ambassadors for Apollo and provided word-of-mouth publicity, which improved the top line.
Another example of decreased turnaround time and a resultant increase in profitability was seen in the Biochemistry
Lab at the Apollo Bangalore Hospital. Dr. Rao headed this lab and he understood the patients’ requirement of receiving diagnostics reports in two hours instead of three. Dr. Rao and his team redesigned the process using 5S and lean concepts and managed to reach a turnaround time of two hours. Profits from the Biochemistry Lab nearly doubled after the decrease in turnaround time; while the cost of consumables increased by only 11%.
Even though the Apollo team was trying to improve customer satisfaction, it still faced the question of how much satisfaction could be actually provided to the customer considering the room tariffs that were charged. As seen from
Exhibit 13, the charges at the Apollo Bangalore Hospital ranged from USD 25 for a basic room to USD 120 for the
Platinum Suite. A Ritz–Carlton basic room would cost USD 799 at Washington, U.S.A. and USD 165 at Kuala
Lumpur, Malaysia (per person, per night). 14 The Apollo team might be able to provide high quality hospitality to patients in the Platinum Suites. However, the aspiration to provide the same service to patients in other rooms might not be financially feasible. The team was trying to build high levels of service for the Platinum Suites. However, the volumes in the other rooms were too high to be ignored, especially in the Indian context. Additionally, customer loyalty was extremely important to Apollo; in Dr. Panyala’s words,
Customer loyalty and not mere retention is what we need to focus on. It is important to think ahead of the customer to identify issues that may compromise the experience.
QUANTIFYING HOSPITALITY ACROSS APOLLO
According to Dr. Rao,
Once, we develop the benchmarks and the Sigma metrics, we want to replicate the system across all Apollo hospitals in the country. Each hospital will have to devise its own benchmark and
Sigma metrics. However, we want to provide a framework for developing these and then measuring the outcomes. All the hospitals would then be compared by equalisation of scores and would benefit from one another’s learning”.
After collecting the feedback and attempting to set benchmarks, Dr. Rao knew that he needed to go deeper and analyze each service through the complaints, set up relevant benchmarks, and target certain Sigma levels for each benchmark. He wondered whether they could collect and analyze data in a better manner. He wanted to arrive at the basis for the cost-benefit analysis of this activity. Looking at the complaints and the analysis, Dr. Rao had two major questions on his mind:
1.
2.
14
What strategy should be used to reduce the number of complaints and sustain the culture of excellence at Apollo Hospitals, Bangalore under the leadership of Dr. Panyala?
Given the manual intensive processes involved in addressing the hospitality issues, what is a good
Sigma level? Could Apollo set a target for Sigma level in hospitality?
Source: www.ritzcarlton.com, accessed on April 15, 2013.
24
Page 7 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 1
ACE@25 parameters
Sl.
Parameter
No.
1 Coronary artery bypass grafting (CABG) mortality rate
2 Complication rate post coronary intervention(percutaneous transluminal coronary angioplasty; PTCA)
3 Average length of stay (ALOS) post angioplasty
4 Average length of stay (ALOS) post total hip replacement (THR)
5 Average length of stay (ALOS) post total knee replacement (TKR)
6 Complication rate for total knee replacement (TKR)
7 Average length of stay (ALOS) post renal transplant
8 Average turnaround per dialysis chair per day
9 Average length of stay (ALOS) post transurethral resection of the prostate(TURP)
10 Complication rate transurethral resection of the prostate(TURP)
11 Endoscopy complication rate
12 Patient satisfaction with pain management
13 Door to thrombolysis time in ischemic stroke in emergency room (ER)
14 Percentage conversion of coronary angiographies to coronary artery bypass grafting (CABG)
15 Catheter-related blood stream infection (CR-BSI)
16 Ventilator associated pneumonia (VAP)
17 Catheter-related urinary tract infection (CR-UTI)
18 Average length of stay (ALOS) in hospital
19 Average length of stay (ALOS) in intensive care unit (ICU)
20 Door to CT time in stroke cases in emergency room (ER)
21 Surgical site infection (SSI – Clean wound)
22 Medication errors
23 Average length of stay (ALOS) post modified radical mastectomy (MRM)
24 Average length of stay (ALOS) post microdisectomy
25 Average urea reduction ratio*
26
Percentage of patients achieving/maintaining haemoglobin level of 11gram or higher after 3 months of dialysis in end stage renal disease (ESRD)
*Optional
25
Page 8 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 2
Accreditation of Apollo Hospitals
Accreditation
Joint Commission International (JCI)
Delhi,
Apollo Hospital Location
Chennai, Hyderabad, Ludhiana,
Bangalore,
Kolkata, Dhaka
National
Accreditation
Board
for
Hospitals
&
Madurai, Chennai
Healthcare Providers (NABH)
National Accreditation Board for Laboratories (NABL)
Chennai
ISO 9002
Chennai
Source: Apollo Investor Presentation (retrieved from www.apollohospitals.com in January 2013)
Exhibit 3
Apollo Brand–Clinical and Hospitality Services
Core Clinical
Services
Hospitality
Source: Interview with Dr. Ananth Rao
26
Page 9 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 4
Feedback Form
YOUR FEEDBACK Thank you for choosing Apollo Hospitals for your healthcare needs. As a quality improvement initiative, we are looking for improvements in parameters towards ‘‘Service Excellence’’ of our hospital.
Please provide a few minutes of your valuable time for a personal interaction.
How satisfied are you with your experience and the services provided by our hospital on a scale of 1 to 10?
1. MEDICAL SERVICES
1
2
3
4
5
6
Poor
2. NURSING SERVICES
1
2
3
7
8
Good
4
5
6
3. OPERATIONS &
ADMINISTRATION
1
2
3
8
5
6
7
1
2
3
1
2
4
3
5
6
7
1
2
3
4
5
Excellent
8
9
10
6
7
Excellent
8
9
Good
4
5
Poor
6
7
8
Good
Patient Name (Optional):
UHID:
Date of Admission:
Signature:
Date:
27
10
Excellent
COMMENTS (OVERALL):
Room No:
10
Good
Poor
6. FACILITY &
MAINTENANCE
9
Good
Poor
5. HOUSEKEEPING
SERVICES
10
Excellent
8
Poor
4. FOOD & BEVERAGES
9
Good
4
10
Excellent
7
Poor
9
9
10
Excellent
Page 10 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 5
Feedback collection methodology
A Typical Process Map: DPA
Daily Data Collection Method
Score given to IT
Dept. & disseminated to individual stakeholders via SMS by 3:00pm
1:00pm–2:30pm
Data
consolidated;
DPA Score developed* 11:00am–1:00pm
Survey conducted
Point of Data
Colle
ction
Details of complaints mailed to individual stakeholders by
3:00pm
th
15 day consolidated
DPA Score calculated
Data submitted to CEO & DMS
Data submitted to CEO & DMS
Patients staying for 1 day or more Discharged patients Monthly
Analysis
*The lower the score, the better
28
Quarterly
Analysis
Sustenance
Audits
Page 11 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 6
Flow for real-time complaint escalation
Survey conducted
Real-time escalation of issues within 15 minutes to stakeholders concerned
Resolution turnaround time (TAT) given by respective stakeholders Closure audits conducted at the end of TAT
Patient surveyed
NO
Issue rectified? YES
Issue closed
Plan preventive action
29
Stakeholder/representative meets patient for better understanding of issue Page 12 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 7
Department-wise complaints
HK
Department
No. of Complaints
373
F&B
318
Facility
236
Operations
189
Nursing
173
Medical
92
Billing
34
Dietary
19
30
Page 13 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 8
Sample complaints from each department in Apollo Hospitals
Housekeeping: Sample Complaints
Housekeeping service needed to be improved.
The dustbin was broken. The patient’s relative went to the housekeeping desk and got a new one, but housekeeping staff came and took the new dustbin away and replaced it with broken dustbin.
Linen was not changed in 24 hours. Today morning, when we requested for fresh linen, they said it would be provided in 5 minutes; it was not changed till 3:15 pm.
The TV remote was not provided in the room. We asked for it 3 to 4 times, but there was no response.
The toilet was stinking. Yesterday evening, the floor was dirty.
Drinking water was not provided when we asked for it, so finally we had to get it from home.
Housekeeping came to clean only once a day. The inpatient guide book said that they would come twice a day.
The clock battery had to be replaced.
Yesterday, I did not get a blanket the entire day.
It took 20 minutes to get a blanket.
Food & Beverages: Sample Complaints
The food served was not hot. I asked for kichidi but got only rice.
Around 8:30 pm last night, the patient asked for milk. We were told it would be served it by 9 pm, but the patient did not get milk. Today, we asked for milk again.
The food served was too spicy and dry. The food served was not what we had ordered.
The food served was not tasty, so we brought food from home.
The quantity of food served was not sufficient. The quality of roti was bad—it was dry and hard.
Coffee and milk were always served cold, even after complaints.
Food quality was not good
It took 1.5–2 hours for the plates taken away.
The cafeteria menu card was spoiled. They should get a new one.
Despite the patient informing the staff that she did not eat food with garlic or onions, the food served contained garlic and onions. After endoscopy was done, it took 1 hour for the food to be served despite repeated reminders.
There was too much pepper in the child’s food.
Facility: Sample Complaints
The nursing call bell did not work properly. The nursing call bell wire was short and did not reach the patient’s bed.
The room was too congested and stuffy. The temperature in the room should be checked regularly.
In the semi-private room, the light from the next room was too bright; the glass partition should be covered with curtains. The TV remote was taken away since the TV disturbed the other patient.
31
Page 14 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 8 (Contd.)
The staff members in the parking area were not cooperative. They were very arrogant. Last time, the patient was admitted to the ER at 11:30pm, but we were not allowed to park even for a few minutes.
The balcony was not clean. When we felt suffocated in the a.c. room, we opened the window. There was stagnant water outside, which led to a lot of mosquitoes coming in.
Hot water was not available in the bathroom.
Extra curtains should be provided. There were hooks for extra curtains. I had asked for curtains, but they were not provided. This hampered my privacy. The door made too much noise, which disturbed my sleep.
The flush was not working properly.
The attendant’s cot size should be a little bigger.
Operations: Sample Complaints
The doctor advised the patient undergo bronchoscopy; and the patient was asked not to eat anything from 8:30 am that day. At 12:30 pm, when the patient’s attender enquired with the bronchoscopy department, the attender was told that the procedure was postponed to the next day. No communication was provided to the patient regarding this; information was given only when the attender specifically asked for it.
The patient’s attender was told that the patient would be shifted to the ward from the OT recovery only after 1 hour, so the patient’s attender went for lunch. However, the patient was shifted in 20minutes.
The insurance staff left at 3:00pm on Sunday and there was nobody to assist us with the formalities.
We came to the Emergency Room at 10am. We were told we would get a room around 2pm. The room was finally allotted at 9pm.
A Telugu translator was required.
A “silence please” board should be put in each room.
The transport boy’s shoes were stinking. The hygiene of shoes should be improved.
All the staff should close the door while leaving the room.
We had to wait for the discharge summary as it was not updated on time.
The patient or the attenders should be made aware of what to bring to the hospital. When we asked the doctor what all we should bring, the doctor told us to get some clothes for the baby. Nobody informed us about the other things that would be required and were not included in the package. Information and orientation should be given beforehand and not at the time of an emergency.
Three days ago, while making an appointment over the phone, the receptionist made me wait for 3–4 minutes the first time. When I called again, the same person made me wait for 7–8 minutes. The appointment was finally given only when I called the third time.
32
Page 15 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 8 (Contd.)
Medical: Sample Complaints
The patient was in severe pain.
We expected the doctor to visit the patient in the morning. He had not come on rounds till 3 pm. We were waiting for the treatment plan and discharge.
Many staff members were seen walking into the canteen wearing OT gowns and shoe covers.
The operation theatre was available earlier, but was not allotted due to coordination issue (the surgery was scheduled at 12pm,but the patient was not shifted even at 3 pm).
The patient was suffering from pain in the stomach.
Backup treatment should be available for oncology patients in the ER and even in the ward.
Nursing: Sample Complaints
The nurses should attend to patients more frequently, especially on Sundays and other holidays.
HDU: there was a slight delay in service. Some of the nursing staff members were not prompt.
Blood sample had already been taken in the morning, but the staff came and took the sample again; they said that the previous blood sample could not be used as the blood had clotted. The sample collection process was extremely painful. Yesterday night, around 11:30pm, the nurse on duty pulled out the IV line in a rough manner, which caused a lot of pain, even though the patient kept asking the nurse to do it slowly.
The response time of nurses was bad. It took them at least half an hour to come after being called 2–3times.
It would be nice to have a specific set of nurses to take care of me regularly.
The nursing staff seemed to be overworked.
A checklist/patient education material for the plan of care should be available for patients’ convenience.
The staff should try to find out the patient's needs. The response time was very high. There should be additional staff, as the staff seemed to be overburdened.
Today morning, the nurse tried thrice but could not find a vein. She then called another nurse, who completed the procedure. My hand hurt for a long time after this.
Billing: Sample Complaints
Billing took a long time. It should be monitored and abnormal delays should be avoided.
During admission, the credit cell department staff behaved very rudely.
The discharge process took more than 2 hours.
What all were included in the package was not very clear to us.
The cost of surgery should be informed beforehand, rather than calling us at short notice and telling us to pay immediately. If we had been informed earlier, it would have been easier for us to arrange the finance and settle it.
We had to wait for 40 minutes for the finalization of the bill.
33
Page 16 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 8 (Contd.)
Dietary: Sample Complaints
Food should be less spicy and should be according to the child's taste.
The dietician did not visit the patient until 4:10pm. The patient was admitted on 9/6; the surgery (knee replacement) was performed on 10/6. The patient would be discharged the next day, and we needed to meet the dietician.
The doctor suggested normal diet but yesterday morning, the dietician changed it to soft diet. In the evening, the doctor changed it to normal diet. However, soft diet was continued until lunch today. After we complained, the F&B boy changed it to normal diet, but I had already finished my lunch by then.
The doctor told us not to give chutney and sambar to the patient. After having the food suggested by the dietician, the patient had stomach pain.
The dietician has not come to my room yet.
Exhibit 9
Text analytics–word frequency count
Sl.
No.
1
Word
Time
Count
271
Sl.
No.
19
Word
Good
Count
64
2
Food
220
20
Response
62
3
Nurse
214
21
Proper
58
4
Clean
206
22
Said
57
5
Toilet
139
23
Hot water
54
6
Water
129
24
Late
48
7
Doctor
107
25
Bedsheet
47
8
Improve
106
26
Quality
46
9
Wait
98
27
Floor
43
10
Staff
95
28
Serve
43
11
Bed
94
29
Lunch
41
12
Morning
94
30
Response time
41
13
Night
87
14
Attend
79
15
Work
79
16
Hot
78
17
HK
69
18
Minute
66
34
Page 17 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 10
SN KPI DEPT
1
F&B
Benchmarks for hospitality developed by Apollo Hospitals, Bangalore
MEASUREMENT
ROUTINE MEAL DELAY
COLOUR
TARGET
CODE
PROCESS
RED
>20MINS
YELLOW 10-20MINS
GREEN <10 MINS
FREQUENCY
FORTNIGHTLY
IN BETWEEN ORDERS DELAY
FOOD
RED
YELLOW
GREEN
>20MINS
5-20MINS
<5 MINS
WEEKLY
IN BETWEEN ORDERS DELAY
BEVERAGES
WRONG DIET
2 DIETARY
POST ADMISSION DIETICIAN
CONSULTATION DELAY
RED
>30MINS
YELLOW 15-30MINS
GREEN 5-15MINS
RED
NA
YELLOW
NA
GREEN
0
RED
DAILY
>120MINS
YELLOW 60-120 MINS
GREEN <60 MINS
3 FACILITY
WEEKLY
DAILY
REPAIR TIME
MINOR
RED
>30 MINS
YELLOW 15-30 MINS
GREEN <15 MINS
WEEKLY
REPAIR TIME
MAJOR
4
HK
5 NURSING
RESPONSE TIME
RESPONSE TIME
RED
YELLOW
GREEN
RED
YELLOW
GREEN
RED
YELLOW
GREEN
35
>60 MINS
30-60 MINS
<30 MINS
11-15 MINS
6-10 MINS
0-5 MINS
>3 MINS
1.5-3 MINS
0.7-1.5 MIN
DAILY
DAILY
DAILY
RED
YELL
KPI
GREEN
OW
SCORE
Page 18 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 11a
Defective/Defects analysis and process re-engineering flowchart
Patient complaint/feedb ack
Identify defect(s) leading to defective Defective
Root
Cause
Analysis
Repeated
Complaints
Process Changes
(Reengineering)
Pilot Study
(PDCA)
If solution acceptable Sustenance Audit
Deploy at all locations 36
If solution not acceptable
Page 19 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 11b
Defective/Defects analysis and process re-engineering flowchart
Delay in food service
Patient did not get food within 20 minutes Detailed
Root
Cause
Analysis
Delay due to various touch points involved
Solutions proposed
Repeated
Complaints
Direct phone line provided to F&B which can be contacted 24*7 from all patient rooms
Manager on duty to follow up Pilot Study
(PDCA) for private rooms
Sustenance
Audit
Deploy at all the wards 37
Solution
Accepted
Page 20 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 12
Exhibit 13
Kano Model
Room tariffs at Apollo Bangalore (as of January 2013)
Bed Type
Charges (USD)
Shared Bed Category-1
25
Shared Bed Category-2
36
Twin Sharing Bed
48
High Dependency Unit (HDU)
48
Private Bed
77
Executive
92
Platinum Suite
120
Intensive Care Unit (ICU)/Critical Care Unit (CCU)
80
38
4188
REV: NOVEMBER 30, 2011
V. KASTURI RANGAN
SUNRU YONG
Soren Chemical:
Why Is the New Swimming Pool Product Sinking?
Jen Moritz grimaced as she reviewed the February 2007 sales report for her company’s new
Coracle product. In September 2006, Soren Chemical had launched Coracle, a new water clarifier for use in small recreational and household swimming pools. Moritz was responsible for marketing the new clarifier, which she believed was a superior product, but the results so far were discouraging.
The volume target was 50,000 gallons (100,000 units) for the first year of sales. However, through the first half of the selling season for pool chemicals, Soren had sold just 3,725 gallons, or 7,450 units.
Moritz also had responsibility for marketing Kailan MW, a clarifier used primarily in large recreational water park facilities with typical capacities of one million gallons or higher. Very small quantities of Kailan MW were sufficient to treat large volumes of water, but it was unsuitable for smaller-scale applications such as residential pools. Thus Soren Chemical had developed Coracle, targeting smaller pools with a lower volume of swimmers (known in the industry as “bather loads”) and a less intense maintenance program.
In 2006, Kailan MW revenues were $6.1 million and sales were on pace for a 7% increase in 2007; it was healthy growth in a relatively mature market. Coracle had been budgeted at $1.5 million in sales for the year, but so far Soren had sold a very disappointing $111,000. The company’s internal analysis estimated a $30 million market in the United States for large-scale, commercial-use clarifiers such as Kailan MW. Moritz was convinced that an even larger market existed for Coracle. She explained her challenge:
How do we convince distributors to push Coracle and retailers to create shelf space for it?
Is the product priced so that the economics are attractive to our channel partners? From the end-user perspective, most pool owners don’t fully understand the safety and cost-saving benefits of Coracle. Should we invest aggressively to create greater awareness?
________________________________________________________________________________________________________________
HBS Professor V. Kasturi Rangan and writer Sunru Yong prepared this case solely as a basis for class discussion and not as an endorsement, a source of primary data, or an illustration of effective or ineffective management. The authors thank Simon Roy of Zodiac Marine & Pool Co. and
Brenda Barnicki of Eastman Chemical Co. for their valuable contributions to the development of this case.
This case, though based on real events, is fictionalized, and any resemblance to actual persons or entities is coincidental. There are occasional references to actual companies in the narration.
Copyright © 2010 Harvard Business School Publishing. To order copies or request permission to reproduce materials, call 1-800-545-7685, write
Harvard Business Publishing, Boston, MA 02163, or go to http://www.hbsp.harvard.edu. No part of this publication may be reproduced, storedin a retrieval system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the permission of Harvard Business Publishing.
Harvard Business Publishing is an affiliate of Harvard Business School.
39
4188 | Soren Chemical: Why Is the New Swimming Pool Product Sinking?
Company Background
Timothy Soren founded the company in 1942 to sell industrial-strength cleaning solutions. In the decades since, the company had expanded its focus to include industrial chemicals for lubricants and fuels, as well as a range of chemical solutions for treating drinking water and wastewater. In 2006, the Soren Chemical product line included over 350 products, and company revenues were $450 million (see Exhibit 1 for financial data).
Historically, Soren Chemical had concentrated on business-to-business sales and placed little emphasis on creating consumer awareness of its products. However, in 2002 the company had begun to invest selectively in developing brands for products that had potential in the consumer market.
While the company had enjoyed only modest success with its branding efforts, Soren Chemical continued to invest opportunistically in products with potential beyond the traditional “B2B” core.
Jen Moritz was a marketing manager in the Water Treatment Products group with responsibility for chemicals used in drinking and pool water treatment. Soren had identified in the pool water clarifier market a significant opportunity to build a consumer brand, and Moritz’s role was to develop the goto-market strategy.
Product Background
Kailan MW and Coracle are “flocculants,” chemicals that cause suspended particles in liquids to agglomerate into larger, heavier particles called “flocs.” They are particularly effective for microscopic particles that cannot be removed through physical filtration alone. Used under the appropriate pH, temperature, and salinity conditions, flocculating agents react with water to form insoluble hydroxides that physically trap small particles into the floc. The larger floc can then be trapped through sedimentation or filtration processes. Moritz explained:
Flocculants can be thought of as chemical “nets” that work in conjunction with filters. To get water clean you need to get rid of particles that would otherwise pass right through filters.
These particles can be as small as 0.5 microns. Only a flocculant can get something so small— the chemical “net” literally traps the little particles that filters miss, and then the net itself gets stopped by the filter.
In 2002, the Soren research and development team created the Kailan MW flocculant and found that it had application as a “clarifier” for pool water. This meant that Kailan MW was effective in dealing with oils that come from deodorant, hair spray, and lotions—which cause odors and cloudiness in water. Cloudy water is a safety hazard, obscuring pool depth for divers and making it difficult for lifeguards to see swimmers below the surface. What set Kailan MW apart from other pool clarifiers was its ability to combat organic debris. It could trap algae as well as dangerous waterborne pathogens such as E. Coli and cryptosporidium—a parasitic disease that normal chlorine levels do not kill. Moritz explained the importance of clarifiers as part of a pool maintenance regime:
The layperson knows only that pools have chlorine and filters. They do not realize there is a complex system of pumps, chemicals, and UV or ozone technology used to make the water safe and clean, with minimal odor. UV systems alone, for example, cannot effectively sanitize water when there is high turbidity or cloudiness. Kailan MW captures the smallest particles and quickly reduces the cloudiness. Similarly, chlorine alone is not sufficient to make water safe because there are resistant, waterborne illnesses such as cryptosporidium. Any large, public pool or water park can be prone to an outbreak. That is where Kailan MW comes in.
2
BRIEFCASES | HARVARD BUSINESS PUBLISHING
40
Soren Chemical: Why Is the New Swimming Pool Product Sinking? | 4188
One gallon of Kailan MW could effectively treat approximately 500,000 gallons of water. Most competing clarifiers required daily application, but Kailan MW’s advantage was its effectiveness over a longer period. Depending on the intensity of pool usage, operators could use it every other day, rather than the daily use required with similar products.
Commercial-Use Clarifiers Market
Soren Chemical estimated the 2007 U.S. market for specialty commercial-use clarifiers to be approximately $30 million, for which primary demand came from commercial pools (there are
300,000 in the U.S.) and water parks (1,000 in the U.S). Most commercial facilities must accommodate high bather loads and a disproportionate number of children, and it is imperative to maintain high levels of clarity and to protect against waterborne pathogens.
Soren Chemical sold Kailan MW for these commercial applications primarily through seven chemical formulators (see Figure A for an overview of the channel structure). The formulators produced or sourced the full range of pool chemicals, including chlorine tablets or liquids, cleansers, enzymes, shock treatments, and algaecides. Most products, including Kailan MW, were sold by the formulators under a private brand name.
Figure A
Pool Chemicals Channel Structure
Chemical Manufacturers
Formulators
Wholesale Distributors
Pool Specialty
Retailers
Mass Retailers
Pool Service
Professionals
Water Parks and
Commercial Pools
Residential Pool Owners
The formulators did more than simply deliver product; they typically provided a range of valueadded services, such as custom blending and packaging services tailored to customer specifications.
In some cases, they developed customized maintenance programs by working with water safety consultants, pump and filter manufacturers, as well as the chemical manufacturers themselves.
These programs included prescribed chemical regimes tailored for different commercial applications based on typical bather loads, filter types, and other cleaning technology. The programs also
HARVARD BUSINESS PUBLISHING | BRIEFCASES
41
3
4188 | Soren Chemical: Why Is the New Swimming Pool Product Sinking?
accounted for regional climate conditions that affected water quality, such as pollen, insects, and pollutants. Moritz explained Soren’s go-to-market strategy with Kailan MW:
These smaller formulators are able to provide custom blends and programs, which often make them preferred suppliers for commercial usage. Soren Chemical’s model does not allow us to customize for individual water parks and commercial pools. Since Kailan MW is bestsuited for this market, it made sense for us to partner with the formulators.
Residential Pool Clarifiers Market
Because Kailan MW was designed for large-scale commercial facilities, Soren Chemical did not intend it for use in smaller pools for fear of misuse and potential safety risks. However, in 2005 the sales team learned that at least two of its formulators had begun marketing a diluted version of
Kailan MW as a private label clarifier for the residential pool market. Moritz recognized that there was a significant, untapped market if the product could be appropriately refined for use in smaller residential pools. After several months of product development, the company unveiled a clarifier featuring chemical structure and properties similar to Kailan MW’s, but designed for smaller pools, lower bather loads, and less-frequent usage. Soren named the new product “Coracle,” after the traditional Welsh fishing boat that moved with the current, sweeping fish into its net.
The potential consumer market for Coracle was more fragmented than the commercial market for
Kailan MW. Industry reports showed that there were nearly 9 million residential swimming pools in the United States, all of which required regular maintenance and cleaning by their owners or a professional pool service. Most pool equipment and chemical manufacturers sold via a two-step process to wholesale distributors, which then supplied specialty retailers and service professionals.
Distributors typically carried tens of thousands of products from many suppliers—including the regional formulators that sold Kailan MW—and served a broad base of local retailers and pool maintenance companies. While there were a few national wholesale distributors, such as Pool
Corporation, many were small regional players. Some manufacturers bypassed the wholesale distributors altogether, selling a limited range of pool products directly to Wal-Mart or “do-ityourself” retailers such as Home Depot and Lowe’s. However, this approach was costly and was used only by the largest chemical companies.
Purchasing behavior differed between professionals and consumers. The professional market, comprising pool builders, cleaning and service companies, and independent contractors, typically bought supplies from the wholesale distributors. Industry experts believed there were 40,000 to
50,000 pool contractors and service companies in the United States. Approximately 80% of consumers maintained their own pools, and they generally purchased supplies from local specialty retailers or national retailers. Whereas the critical concerns with cloudy water in the commercial market were waterborne disease and swimmer safety, the consumer market emphasized aesthetics and perceived cleanliness.
Moritz identified three leading competitors for residential pool-use clarifiers: Keystone Chemicals,
Kymera, and Jackson Laboratories. She estimated that each of the companies had a 15% to 20% share of the residential pool clarifier market. Based on discussions with multiple distributors, she pieced together a product and pricing comparison (see Table A).
4
BRIEFCASES | HARVARD BUSINESS PUBLISHING
42
Soren Chemical: Why Is the New Swimming Pool Product Sinking? | 4188
Table A
Product Comparison
Product
Soren
Chemical
Coracle
Cost per container, retail price
Form
Gallons per container
Cost per gallon, retail price
Cost per ounce/tablet
Ounces/tablets needed per 30K gallons
Cost per treatment, retail price
Replacement cycle (no. of monthly treatments)
Monthly cost, retail price
Average months of pool usage
Annual cost, retail price
Keystone
Chemicals
Purity
Kymera
HydroPill
Jackson
Labs
ClearBlu
$25.00
Liquid
0.50
$50.00
$0.39
10.0
$15.00
Liquid
0.25
$60.00
$0.47
5.0
$3.50
Tablet
NA
NA
$3.50
1.0
$15.00
Liquid
1.00
$15.00
$0.12
32.0
$3.91
2.0
$7.81
5.0
$2.34
4.0
$9.38
5.0
$3.50
2.0
$7.00
5.0
$3.75
4.0
$15.00
5.0
$39.06
$46.88
$35.00
$75.00
Note: Costs are rounded.
Most service professionals considered ClearBlu to be the most effective chemical flocculant.
Neither Keystone’s Purity nor Kymera’s HydroPill had a discernable effect in reducing the need for chlorine, shock treatments, or enzymes. In contrast, industry experts believed that ClearBlu reduced the need for other chemicals by approximately 15%, although Jackson Laboratories did not emphasize this benefit. The key disadvantage of ClearBlu was the relatively high dosage required per treatment: a quarter-gallon of product was required each week. This was higher than the dosage required with the more concentrated products of competitors, making it less convenient for pool owners and service professionals to store ClearBlu in bulk.
Marketing Strategy for Coracle
A study by Soren Chemical showed that Coracle, like Jackson Laboratories’ ClearBlu, significantly boosted the efficiency of other pool chemicals by reducing the “burden” placed on chlorine. Moritz described the benefits of Coracle:
In a typical pool, two-thirds or more of chlorine is used up oxidizing organic contaminants such as oils. This keeps the chlorine from sanitizing the water, which really defeats its primary purpose. Coracle takes care of these oils, freeing up the chlorine to serve its primary function of fighting waste. It also reduces the need for shock treatments and enzymes.1
The potential cost savings for pool owners could be significant. The company’s research and development team estimated that Coracle reduced the need for additional chlorine, shock treatments,
1 Shock treatments or shock oxidizers are heavy doses of chemicals (typically chlorine) used in conventional pools to clean out
contaminants. Shock treatments may be required once a week or more, depending on the intensity of pool usage. Due to the high dosage of chemicals, shock treatments can make pools unusable for an hour or more. Enzymes help reduce organics and phosphates in water and are primarily intended to reduce “scum-line” formation at the edge of the pool, making it easier to clean. HARVARD BUSINESS PUBLISHING | BRIEFCASES
43
5
4188 | Soren Chemical: Why Is the New Swimming Pool Product Sinking?
and enzymes, reducing pool owners’ annual chemical costs by 20% to 30%.2 Soren Chemical made this benefit the central thrust of its marketing message, prominently featuring the claim on the bottle.
In order to reach service professionals and specialty retailers, Soren sold Coracle through wholesale distributors. The company allowed formulators to use private branding for Kailan MW and, indeed, most opted to sell clarifiers that did not explicitly identify Kailan MW as the flocculant.
However, the company decided not to allow private-label branding for Coracle, despite requests from several major wholesale distributors that marketed a number of pool chemicals under their own labels. Moritz explained how this fit into a longer-term plan:
In keeping with Soren’s new strategy to opportunistically develop consumer brands, our plan with Coracle is to make it a branded product. Our R&D pipeline includes other products, such as algae and phosphate removers that have potential in residential pools and spas. If we can build some recognition with the Coracle name, we have a platform for other products.
Ideally, Coracle will become almost a consumer packaged good that pool service professionals, specialty retailers, and consumers will know and request.
The manufacturer price for a two-quart bottle of Coracle was $14.88. For most products such as chlorine, shock treatments, and enzymes, distributors typically had a 20% gross margin. However, wholesale distributors expected to maintain a 30% gross margin in selling Coracle, as it was a differentiated chemical agent. Retailers and service professionals usually took a 15% gross margin, resulting in a suggested retail price of $25 to consumers (see Exhibit 2 for margin structure data).
The product launch in fall of 2006 was accompanied by a new website and a press release in three trade association journals targeted at pool service professionals and specialty retailers. The announcement emphasized Coracle’s performance advantage in trapping dangerous waterborne pathogens such as E. Coli and cryptosporidium. Soren Chemical received over 2,000 inquiries from service professionals and retailers in the first three months of 2007. The company responded directly with a brochure, technical notes, and material safety data sheet.
Despite these efforts, sales through February were a dismal $111,000.
The Dilemma
Frustrated with such underperformance, Moritz reviewed her marketing plan with colleagues, noting that Soren Chemical’s relative inexperience with marketing consumer-oriented brands appeared to be hindering the efforts with Coracle. One colleague questioned whether Moritz had fully accounted for competing products:
We are new to the residential pool market and perhaps we do not fully understand the consumer. How is Coracle beneficial to them and do they understand that? They already have other choices, and Coracle is around $25, which is higher than the unit prices of competitors. I think we need to use the pricing of comparable products as our guideline.
A second colleague pointed out the challenges of selling through distributors and retailers:
2 Cost savings estimates are based on chemical usage for conventional pools, which comprise over 85% of total swimming pools in the U.S. Salt water pools use significantly less chemicals than conventional pools, although clarifiers provide similar water safety benefits.
6
BRIEFCASES | HARVARD BUSINESS PUBLISHING
44
Soren Chemical: Why Is the New Swimming Pool Product Sinking? | 4188
We need to be certain the economics are attractive for distributors and retailers. After all, they already carry other clarifiers. Why would they make shelf space to carry Coracle? I worry that their incentives are not totally aligned with ours. If we have room to raise prices, it might create enough margin to make this more attractive to them.
To diagnose the situation more carefully, Moritz decided to conduct a survey among the pool service professionals and specialty retailers who had made inquiries about Coracle. The survey revealed that only 30% of the respondents recalled receiving the Coracle materials that Soren
Chemical had sent in response to their inquiries. Furthermore, though Soren Chemical had passed the contact information for interested customers to the appropriate wholesale distributors, nearly
70% of respondents stated that Coracle had not been offered by their distributors. Creating greater demand from service professionals could provide a jumpstart for Coracle, but Moritz estimated that she would need a $600,000 budget to conduct a mailing campaign and to run advertising in industry publications. Moritz also used the survey to learn from retailers and professionals about consumers’ use and understanding of clarifiers (see Exhibit 3). Soren Chemical had priced Coracle aggressively at a $25 suggested retail price, and the estimated $39 annual cost for the average pool was lower than with most competing products. Moritz suspected that most residential pool owners simply did not realize the value of Coracle relative to other clarifiers.
The selling season for residential pool chemicals would be largely over by May. Moritz needed to reevaluate the entire go-to-market strategy. If she could identify the problem, perhaps she could partly salvage 2007 and position Coracle for a successful second year.
HARVARD BUSINESS PUBLISHING | BRIEFCASES
45
7
4188 | Soren Chemical: Why Is the New Swimming Pool Product Sinking?
Exhibit 1
Soren Financial Data ($ millions)
2004
2005
2006
$ 439
335
104
$ 444
339
105
$ 450
345
105
Selling, General, & Administrative Expense
Research & Development
Depreciation
29
8
15
30
8
16
31
9
15
Operating Expense
52
54
55
Operating Income
52
51
50
Provision for Tax
Net Income
12
40
11
40
12
38
121
36
$ 15
130
39
$ 17
134
40
$ 17
Cost per
Container
Cost per
Treatment
Revenue
Cost of Goods Sold
Gross Profit
Water Treatment Group
Revenue
Gross Profit
Operating Income
Exhibit 2
Margin Structure
Retailer / Service Professional price
Retailer / Service Professional gross margin %
Retailer / Service Professional gross profit
$25.00
15%
$3.75
$3.91
15%
$0.59
Distributor price
Distributor gross margin %
Distributor gross profit
$21.25
30%
$6.38
$3.32
30%
$1.00
Soren Chemical price
Soren Chemical gross margin %
Soren Chemical gross profit
$14.88
35%
$5.21
$2.32
35%
$0.81
Exhibit 3
Specialty Retailer Survey, Selected Data
Annual chemical costs at retail prices, excluding clarifiers (recommended regime)
% of consumers who understand and use clarifiers regularly
Annual average cost of clarifiers at retail prices (recommended regime)
8
$300 average
25%
$50
BRIEFCASES | HARVARD BUSINESS PUBLISHING
46
Harvard Business School
9-599-114
May 25, 1999
Note on Behavioral Pricing
It is important for a firm to get its pricing “right.” Consider the impact of product pricing on a firm’s net income. If Coca-Cola could increase its prices by an average of 1%, without affecting consumer demand for its products, it would increase its net income by 6.4%. A price increase of less than 1¢ on a can of cola would translate to an increase in net income of about $300 million.
Similar 1% price increases, if they did not negatively impact demand, would lead to increases in net income of 16.7% for Fuji Photo, 17.5% for Nestle, and 26% for the Ford Motor Company. In fact, an average price increase of 1% would boost the net income of the typical large U. S. corporation by about 12%.1
These examples highlight the potential impact on a firm’s net income of optimally setting product prices — small changes in price can have an enormous impact on income. Before raising (or lowering) prices in an attempt to improve the bottom line, however, a firm must understand and anticipate a consumer’s response to a product price change.
Unfortunately, it appears that many firms lack the necessary understanding of a consumer’s
“willingness to pay” to optimally set product prices. When asked whether they were “wellinformed” on six of the potential inputs to the product pricing decision, managers at one wellrespected U.S.-based multinational responded as follows:2
•
•
•
•
•
•
84%
81%
75%
61%
34%
21%
were well-informed on the variable cost of providing their product. were well-informed on the fixed cost of providing their product. were well-informed on the price of competitors’ products. were well-informed on the value of their product to the customer. were well-informed on how consumers would respond to price changes. were well-informed on consumers’ willingness to pay at various price levels.
These managers were well-informed on the costs of providing its products and on the price of competitor’s products. They were also well-informed on the value its products delivered to consumers. However, when it came to a consumer’s willingness to pay or to a consumer’s response
1
2
From Robert J. Dolan and Hermann Simon’s, Power Pricing, The Free Press, New York, NY (1995)
Ibid.
Assistant Professor John T. Gourville prepared this note as the basis for class discussion rather than to illustrate either effective or ineffective handling of an administrative situation.
Copyright © 1999 by the President and Fellows of Harvard College. To order copies or request permission to reproduce materials, call 1-800-545-7685, write Harvard Business School Publishing, Boston, MA 02163, or go to http://www.hbsp.harvard.edu. No part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the permission of Harvard Business School.
1
47
599-114
Note on Behavioral Pricing
to potential price changes, these managers were lacking the insight needed to optimally set prices.
Experience suggests that this company is not unique in this regard.
Purpose of this Note
Any firm’s ability to optimally set product prices is governed by many factors. Some of these factors are well understood and are routinely incorporated into a firm’s pricing decisions. These factors tend to be heavily weighted toward those economic factors that are easily obtained by a firm, including their own variable and fixed costs of production, the market price of competitors’ products and the firm’s internal assessment of the value that their product delivers to the intended consumer.
Not surprisingly, these are the four factors on which the surveyed managers claimed that they were well-informed. We will briefly review these factors in a moment. 3
However, optimal product pricing also hinges on a consumer’s willingness to pay and on a consumer’s response to price changes, factors on which the surveyed managers claimed they were poorly informed. In addition, research has shown that a consumer’s willingness to pay is often influenced by “psychological” or “behavioral” variables that typically are not considered when setting price. Specifically, consumers often are as concerned with the behavioral question of “how fair a deal am I getting” as they are with the economic question of “how good a deal am I getting.”
This note is an attempt to highlight the potential impact of some of these behavioral variables on a consumer’s willingness to pay. In the process, it provides a more complete picture of consumer response to pricing and provides some insight for optimal product pricing.
Value Pricing and the Economic Perspective
The traditional economic approach to product pricing is driven by a small handful of factors, as shown in Figure A. One of these factors is the “objective value” the product delivers to the consumer. 4 This is a measure of the benefits that the product delivers to the consumer, regardless of whether the consumer recognizes those benefits. When 61% of managers claim they are wellinformed on the value of their product to the consumer, they are most likely referring to this
“objective value.”
A second factor in the economic approach to pricing is the “perceived value” of the product to a consumer. Perceived value is the value the consumer understands the product to deliver.
Sometimes, a product’s benefits are readily apparent to the consumer and “perceived value” approaches “objective value” with little effort by the firm. Other times, a product’s benefits are less obvious and need to be communicated by the firm to the consumer (e.g., via advertising, personnel selling). In such cases, the “perceived value” of a product typically falls below its “objective value.”
The perceived value of a product also can be influenced by the price of competing products or “substitutes.” Company A may develop a product that creates great objective value for consumers.
Consumers may recognize this value and be willing to pay a high price to obtain the product.
3
For a more complete discussion of these economic factors, you should refer to Professor Corey’s “Note on
Pricing” [HBS Note #580-091] or Professor Dolan’s “Pricing Policy” note [HBS Note #585-044].
4
Given consumer heterogeneity, “objective value” and “perceived value” will tend to vary across consumers.
For some consumers, these values will be high, for others, they will be low or zero. For simplicity, we will ignore consumer heterogeneity in this note and only consider the “typical” consumer.
Nevertheless, the behavioral perspective offered in this note apply equally well to a heterogeneous consumer population.
2
48
Note on Behavioral Pricing
599-114
However, if Company B introduces an identical product at a much lower price, the perceived value of Company A’s product would be reduced to the price of Company B’s product.
It is important to note that the “perceived value” of a product to a consumer should equal the maximum price that consumer is willing to pay for the product. Imagine a consumer who perceives the value of a modem to be $100. If priced above $100, the consumer has no incentive to buy the modem. If priced at $100 or less, however, the consumer always stands to gain from purchasing.
Figure A
Value Pricing and the Economic Perspective
Marketing
Efforts
Objective Value
Perceived Value
Consumer’s Incentive to Purchase
= [Perceived Value - Price]
Price of
Substitutes
Product Price
Firm’s Incentive to Sell
= [Price- COGS]
Cost of Goods Sold
$0
The last major component to the economic approach to pricing involves the firm’s cost of goods sold (i.e., COGS). Just as the consumer requires an incentive to purchase a product, the firm requires an incentive to sell the product. In order to stay in business and make a positive return, a firm must charge a price that covers both its cost of production. 5
All of these economic factors come together to form the “value pricing” approach to pricing.
In optimally pricing a product, a firm is bound at the upper end by the consumers’ “perceived value” for the product. This “perceived value” is influenced by the “objective value” of the product to the consumer, by the firm’s marketing effort to communicate that objective value, and by the price of substitute products. At the same time, the firm is bound on the lower end by its COGS.
By pricing above COGS and below perceived value, the firm has an incentive to sell the product, measured as [price – COGS], and the consumer has an incentive to purchase the product, measured as [perceived value – price]. In value pricing terminology, the firm has “created” value by offering a product that the consumer values at a price greater than the firm’s COGS. In turn, by pricing between perceived value and COGS, the firm has “captured” some of that value for itself and has allowed consumers to capture the remainder.
5
For simplicity, we will ignore strategic reasons for pricing below cost such as to build share or volume or to temporarily respond to a competitor’s pricing efforts.
3
49
599-114
Note on Behavioral Pricing
Adding a Behavioral Component to the Economic Perspective
This “value pricing” framework provides a basic model of how an economically rational consumer should respond to a firm’s pricing of a product. A rational consumer should purchase a product as long as the “perceived value” of that product is greater than the actual price being charged. In addition, the more one’s perceived value exceeds actual price, the greater should be a consumer’s incentive to buy. This leads to the fairly straight forward claim that:
The Economic Perspective: Consumers Buy When Perceived Value Exceeds Price
Consumers should purchase an item whenever the perceived value of that item exceeds its actual price [i.e., whenever (Perceived Value – Actual Price) > 0].
Adding a Behavioral Component
To complement this economic perspective, we now add a “behavioral” or “psychological” perspective to product pricing. This perspective captures “how fair a deal” one is getting. To make this point clear, consider the following scenarios first proposed by Professor Richard Thaler. 6
Scenario #1:
You are lying on the beach on a hot day. All you have to drink is ice water. For the past hour, you have been thinking about how much you would enjoy a nice cold bottle of your favorite beer. A friend gets up to make a phone call and offers to bring back a bottle of your favorite beer from the only nearby place where beer is sold — a small, run down-grocery store.
He says that the beer might be expensive and asks how much you are willing to spend. He says he will not buy the beer if it costs more than the price you state. What price do you tell your friend?
What is your “perceived value” for a nice cold bottle of your favorite beer brought to you on a hot beach? Once you have decided upon the price you would tell your friend, consider a second scenario, identical to the first, except for the source of the beer, which is underlined.
Scenario #2:
You are lying on the beach on a hot day. All you have to drink is ice water. For the past hour, you have been thinking about how much you would enjoy a nice cold bottle of your favorite beer. A friend gets up to make a phone call and offers to bring back a bottle of your favorite beer from the only nearby place where beer is sold — a fancy resort hotel. He says that the beer might be expensive and asks how much you are willing to spend. He says he will not buy the beer if it costs more than the price you state. What price do you tell your friend?
If you are like most people, your responses to these two scenarios differ. When Thaler presented these scenarios to a group of executives in the early 1980s, the median response in the grocery store scenario was $1.50 and the median response in the fancy resort hotel scenario was $2.65.
Similar results repeatedly have been obtained using Harvard MBAs — albeit, with somewhat higher average prices.
6
Scenarios 1 and 2 have been adapted from Richard H. Thaler’s paper, “Mental Accounting and Consumer
Choice,” Marketing Science, 4, 3 (Summer 1985): p. 199-214.
4
50
Note on Behavioral Pricing
599-114
The interesting question is why? As pointed out by Thaler, for the person consuming the beer on the beach, nothing of importance has changed between the two scenarios. Specifically,
•
in both scenarios, the ultimate consumption is identical — the same beer is consumed on the same beach.
•
no atmosphere from the fancy resort hotel or the run-down grocery store is being consumed by the beer drinker to justify different prices.
•
there is no strategic reason to report a price below one’s “perceived value” for the beer [e.g., you cannot haggle over price with hotel or store owner].
As a result, a person’s “perceived value” for the beer should be identical across the two scenarios. To report otherwise would suggest that a bottle of beer consumed on a beach somehow tastes better or quenches thirst more effectively when purchased from one location than another. In turn, if the “perceived value” of the beer should be identical across the two scenarios, a person’s
“willingness to pay” also should be identical across the two scenarios.
Yet people’s prices do differ and it appears that this difference is due to expectations consumers have regarding the price of a bottle of beer at a fancy hotel versus a run-down grocery store. As noted by Thaler, “While paying $2.50 for a beer is an expected annoyance at the resort hotel, it would be considered an outrageous ‘rip-off’ in a grocery store.”
In the end, it appears that one’s “willingness to pay” in these two scenarios is driven not only by the “economic utility” of the transaction [i.e., perceived value – price], but also by the
“psychological utility” of the transaction, driven largely by a consumer’s perception of “fairness.”
Over the next several pages, we will look at scenarios that highlight specific drivers of transaction
“fairness” and of the “psychological utility” of a transaction.
Some Behavioral Updates to the Economic Perspective
1.
The Relative versus Absolute Value of Money
In the economic approach to pricing, all money is equal – e.g., $10 in one transaction is worth the same as $10 in another transaction. Research suggests that this may not be the case when it comes to one’s willingness to pay, however. Consider the following scenario and think about how you would respond. 7 There are no right or wrong answers. There is only your intuition as to how you would behave if you found yourself in such a scenario.
Scenario #3:
You set off to buy a Sony Walkman at what you believe to be the cheapest store in the area.
Upon arriving, you find that the Walkman you want costs $29, a price consistent with your prior expectations. As you are about to make the purchase, a reliable friend tells you that the very same Walkman is selling for $10 less at a store approximately 10 minutes away. Do you go to the other store to buy the Walkman?
7
Scenarios 3 and 4 have been adapted from Richard H. Thaler’s paper, “Toward a Positive Theory of
Consumer Choice,” Journal of Economic Behavior and Organization, 1 (1980), p. 39-60.
5
51
599-114
Note on Behavioral Pricing
What would you do? Do you go to the other store and save $10 or do you simply go ahead and purchase the Walkman in the first store and pay $29?
A purely economic approach to this question would be to ask yourself whether 10 minutes of your time is worth $10. If you decide that 10 minutes of your time is more valuable than $10, you should forget about the potential savings and purchase the Walkman at the first store for $29. If, however, you decide that 10 minutes of your time is less valuable than $10, you should travel to the second store and purchase the desired Walkman there for $19.
Now consider a second scenario, identical to the first, except for the nature and the price of the product being purchased.
Scenario #4:
You set off to buy a Sony Camcorder at what you believe to be the cheapest store in the area.
Upon arriving, you find that the Camcorder you want costs $495, a price consistent with your prior expectations. As you are about to make the purchase, a reliable friend tells you that the very same Walkman is selling for $10 less at a store approximately 10 minutes away.
Do you go to the other store to buy the Camcorder?
From the economic perspective, if you decided to travel to the second store to save $10 in the first scenario, you also should have decided to travel to the second store to save $10 in this second scenario. In both scenarios, the tradeoff is $10 for 10 minutes of your time.
If you are like most people, however, your natural inclination will be to answer “yes” in
Scenario #3 and “no” in Scenario #4. After all, $10 on a $29 Walkman represents a savings of over
33%, but $10 on a $495 Camcorder represents a savings of a measly 2%. While the dollar savings are the same, the psychological value of the savings is far greater in the first scenario than the second.
Whereas a $10 savings on a $29 Walkman is perceived as a “fair” (and even generous) incentive to travel to the second store, a $10 savings on a $495 camcorder is perceived as a rather inadequate incentive to travel to the second store.
These two scenarios raise a curious and important fact about money. Namely, the
“psychological utility” of a fixed amount of money (e.g., $10) is relative. Saving $10 on a $29 item will have much greater impact on a consumer’s behavior than saving $10 on a $495 item.
This directly carries over to a consumer’s willingness to pay. Imagine two consumers, one of whom is debating whether to pay $19 for a Walkman that he values at $29, the other of whom is debating whether to pay $485 for a camcorder that he values at $495. While both consumers have the same “economic utility” to enter their respective transactions [i.e., perceived value – price = $10 in both cases], the first consumer will be more likely to make a purchase than the second due to the higher relative incentive to enter his transaction.
This leads to our first update to the economic perspective:
Behavioral Update #1: Willingness to Pay is Impacted by Relative Incentives
In determining his willingness to pay, a consumer will consider both his absolute
“economic utility” from the transaction [i.e., perceived value – actual price] and his relative incentive to enter the transaction [i.e., (perceived value – actual price)/(actual price)].
6
52
Note on Behavioral Pricing
599-114
2. The Impact of a Salient Reference Price
Expectations about “what a product will cost” also seem to impact a consumer’s “willingness to pay.” 8 Read the following scenario and think about how you would respond. Again, there are no right or wrong answers. Only your intuition matters.
Scenario #5
Your favorite sports team has made the playoffs. Its first-round playoff series is a best-ofseven series9 with Games 1, 2, 5, and 7 played on your team’s home field. General admission tickets had been priced at $20 during the regular season. The team decided to raise general admission prices to $40 for these four playoff games. Is this price increase fair or unfair?
Is it fair for a sports team to raise prices by $20 between the regular season and the playoffs?
Most people who encounter this scenario say “yes.” About two-thirds of Harvard MBAs not only believe it to be “fair,” but would expect such an increase. Arguments in support of this stance include: •
The nature of the product has changed – playoff games are more exciting than regular season games.
•
Demand almost certainly will be higher for the playoff games and the team is merely responding to this increase in demand.
•
It is common for ticket prices to increase for playoff games.
Now consider a second scenario, similar to the first scenario except for the timing of the $20 price increase.
Scenario #6
Your favorite sports team has made the playoffs. Its first-round playoff series is a best-ofseven series with Games 1, 2, 5, and 7 played on your team’s home field. General admission tickets had been priced at $20 during the regular season. General admission tickets were also priced at $20 for Games 1 and 2 of the playoffs. After Game 2, the team decided to raise prices to $40 for Games 5 and 7. Is this price increase fair or unfair?
In considering this second scenario, note that from a purely economic perspective, if Games 5 and 7 are worth $40 in Scenario #5, they should also be worth $40 in Scenario #6. Therefore, any difference in “fairness” between the two scenarios is not being driven by a change in the “economic utility” of the transaction, as measured by [perceived value – price], but by the “psychological utility” of the transaction.
8
In these and subsequent scenarios, individuals are being asked to assess the “fairness” of a firm’s pricing decision. Perceptions of fairness should impact “willingness to pay” in a straightforward fashion. In particular, in the short-term, consumers will be less willing to pay a price they feel is “unfair.” In the long-term, consumers will be less likely to purchase from a firm that makes an unfair pricing decision.
9
In a best-of-seven series, the teams play games until one team wins four. This guarantees a minimum of four games and a maximum of seven. Often, Games 1, 2, 5, and 7 are played at one teams location, Games 3, 4 and 6 are played at the other teams location. Also, for each of Games 5, 6 and 7, tickets often are not sold until it is apparent that each game will be required.
7
53
599-114
Note on Behavioral Pricing
If you are like most individuals, however, you find the price increase in Scenario #6 to be
“unfair.” What seems to drive this sense of unfairness is the fact that ticket prices were raised midplayoffs. Once ticket prices were set for Games 1 and 2, those prices created an expectation for prices for the remainder of the playoffs. By raising prices for Games 5 and 7, this expectation was violated.
These two scenarios highlight the power of a salient “reference price,” a price against which consumers compare other prices to assess both the “goodness” and “fairness” of a given transaction.
In Scenario #5, regular season ticket prices were not a salient (nor appropriate) reference price due to the fundamental difference between regular season games and playoff games. In contrast, in
Scenario #6, ticket prices for playoff Games 1 and 2 did establish a salient (and seemingly appropriate) reference price for Games 5 and 7. As a result, price changes in Scenario #5 are deemed
“fair” while price changes in Scenario #6 are deemed “unfair.”
Interestingly, Scenario #6 was actually played out in the spring of 1997. In a best-of-seven playoff series between basketball’s Miami Heat and New York Knicks, Miami raised ticket prices in the middle of the playoff series. For Games 1 and 2, Miami Heat management had set the prices for various seats at $20, $30 and $40. After Game 2, the Heat raised ticket prices to $50, $80 and $90 for
Game 5. Public outrage resulted and Game 5 was one of the very few basketball playoff games that year not to sell out. The extent of the outrage forced Miami Heat management to return prices to $20,
$30 and $40 for the final and deciding Game 7.
In more mainstream consumer transactions, how are reference prices formed? The most common basis for a reference price is the previous price paid for a product. If a consumer has been paying $9.99 for bottle of wine that they have come to enjoy, this $9.99 price becomes this person’s reference price for the wine. If the consumer subsequently encounters the same bottle of wine at the same store for $14.99, the price difference will likely be questioned — not because the wine isn’t worth $14.99 to the consumer, but because the consumer has grown to view $9.99 as the “fair price.”
This concept is captured in our second update to the economic perspective:
Behavioral Update #2: Willingness to Pay is Impacted by Salient Reference Prices
In determining her willingness to pay, a consumer will consider her “economic utility” from the transaction [i.e., perceived value – actual price] and the consistency between the actual price and a salient reference price [i.e., actual price – reference price].
3.
The Impact of a Firm’s Cost of Goods Sold
Scenarios #5 and #6 show that a salient reference price can impact a consumer’s perception of fairness (and, by extension, her willingness to pay). Another factor that impacts a consumer’s perception of “fairness” is a firm’s cost of goods sold, as indicated by the following two scenarios.10
Scenario #7:
A grocery store has no peanut butter in stock, but is about to receive a new shipment. Prior to delivery, the owner finds out that the wholesale price of peanut butter has increased 20% and will affect this new shipment. The owner decides to increase the price of the new peanut butter by 20%. Is this retailer’s actions fair or unfair?
10 These two scenarios are adapted from Kahneman, Knetsch and Thaler, “Fairness as a Constraint on Profit
Seeking: Entitlements in the Market,” American Economic Review 76, 4 (1986): p. 728-741.
8
54
Note on Behavioral Pricing
599-114
With little argument, most consumers find this retailer’s actions to be entirely “fair.” After all, the retailer is only passing along a wholesale price increase to the consumer. The retailer has not caused the price increase and is not benefiting from it. Contrast that with the following scenario:
Scenario #8:
A grocery store has a one week supply of peanut butter in stock and is due to receive a new shipment in the near future. Prior to delivery, the owner finds out that the wholesale price of peanut butter has increased 20% and will affect the new shipment. The owner decides to immediately increase the shelf price on his current stock of peanut butter by 20%. Is this retailer’s actions fair or unfair?
In thinking about these two scenarios, let us again consider the purely economic perspective.
In particular, one’s “perceived value” for peanut butter should not differ between the two scenarios – the peanut butter is the same in both scenarios. In addition, the price being charged for the peanut butter has increased 20% in both cases. Therefore, any sense of “fairness” is not being driven by the
“economic utility” of the transaction [i.e., perceived value – actual price], but by the “psychological utility” of the transaction.
Nonetheless, most consumers find the retailer’s actions “unfair” in Scenario #8. Why? The typical argument is that this retailer is only entitled to pass on a cost increase on a product that has been subject to that increase. In Scenario #8, by raising the price of the in-stock peanut butter that has not been subject to the wholesale price increase, the retailer is “taking advantage of the consumer.”
In addition to the magnitude of their own incentive to purchase, these two scenarios suggest that consumers are concerned with the magnitude of the firm’s incentive to sell [i.e., actual price – cost of goods sold]. In particular, consumers do not want to be taken advantage of and desire a fair division of value between themselves and the firm that makes a product. As a result, they are willing to label a price increase “unfair” when that price increase is coupled with little or no change in the cost of goods sold. This concept is captured in our third update to the economic perspective.
Behavioral Update #3: Willingness to Pay is Impacted by Cost of Goods Sold
In determining his willingness to pay, a consumer will consider his own
“economic utility” from the transaction [i.e., perceived value – actual price] and that of the firm [i.e., actual price – cost of goods sold].
4.
The Nature of the Product Being Sold
To this point, it has been shown that factors such as a salient reference price or a known cost of goods sold can impact perceptions of transaction fairness. How do these perceptions of fairness vary by the type of product being sold? Consider the following scenario:
Scenario #9
In 1996, baseball’s Seattle Mariners made it to the American League playoffs. During the season, general admission to a Mariners game cost $15. For the playoffs, the Mariners raised the price of general admission tickets to $20. Is this fair or unfair?
Most individuals who encounter this scenario feel that this price increase is fair and justified, for many of the same reasons that a ticket price increase in Scenario #5 was fair and justified. In particular, the underlying conditions have changed (playoffs vs. regular season) and consumer
9
55
599-114
Note on Behavioral Pricing
demand almost certainly will increase. Under those conditions, a price increase is reasonable and should be expected. Now contrast this first scenario with the following:
Scenario #10
A hardware store had been selling snow shovels for $15. The morning after a large snowstorm, the store raises the price of its snow shovels to $20. Is this fair or unfair? 11
From an economic perspective, the arguments that support a price increase in Scenario #9 also seem to apply in Scenario #10. In this second scenario, the underlying conditions have changed
(snow vs. no snow) and demand almost certainly will increase.
Nonetheless, while the price increase in the first scenario generally is viewed as “fair,” the price increase in the second scenario is almost universally viewed as “unfair.” Why the difference?
One possible explanation lies in the nature of the product being promoted. Whereas the purchase of tickets to a Seattle Mariners playoff game is viewed as a discretionary expense, a purchase of a snow shovel after a large snowstorm is viewed as a necessary expense. As a result, to raise the price of tickets when your team makes it to the playoffs is viewed as legitimate, but to raise the price of snow shovels after a snowstorm is viewed as exploitation.
These final two scenarios provide our last update to the economic model.
Behavioral Update #4:
Perceptions of Fairness Vary Across Product Categories
In determining her willingness to pay, the degree to which a consumer will rely upon her “economic utility” from the transaction [i.e., perceived value – actual price] will vary across product categories (e.g., discretionary vs. necessity, luxury vs. utilitarian).
Managing Perceptions of Transaction Fairness
This note was designed to highlight some of the psychological drivers of consumer price response. As captured in Figure B, it adds a behavioral component to the more familiar economic approach to product pricing.
Figure B
Combining the Economic and Behavioral Drivers of Willingness to Pay
Consumer
Willingness to Pay
=
Economic Utility of the Transaction
[Perceived Value – Actual Price]
+
Fairness of the
Transaction
[(Perceived Value – Actual Price)/(Actual Price)]
[Actual Price – Expected or Reference Price]
[Actual Price – Cost of Goods Sold]
11 Scenario #10 is adapted from Kahneman, Knetsch and Thaler, “Fairness as a Constraint on Profit Seeking:
Entitlements in the Market,” American Economic Review 76, 4 (1986): p. 728-741.
10
56
Note on Behavioral Pricing
599-114
This framework suggests that an economically rational consumer should decide whether to buy a product solely on the economic utility of the transaction — i.e., by comparing perceived value to product price. In reality, consumers incorporate a host of psychological or behavioral factors into their decision making. These factors include the relative size of the incentive to purchase, the consistency between actual price and expected price, and the difference between product price and the firm’s cost of goods sold. These behavioral factors influence perceptions of transaction “fairness” and tend to reduce a consumer’s willingness to pay relative to a purely economic perspective.
Armed with these insights, however, how can a firm more effectively manage perceptions of transaction fairness and anticipate a consumer’s willingness to pay for its products. Two strategies are recommended.
Strategy #1: Actively Manage Price Expectations
A consumer usually enters a transaction with some expectation about the price of a product.
As evidenced by the scenarios presented in this note, these expectations can have systematic and significant effects on that consumer’s willingness to pay for the product. Unfortunately, firms often do very little to understand and/or manage these expectations.
Instead, firms should actively manage the reference prices and comparisons that consumers employ when assessing product prices. In particular, a firm should look to:
•
Establish credible reference prices.
A firm should look to establish a benchmark price for its products whenever possible. Common tools for doing this include the use of a credible “suggested retail price” or “list price.” Through the clear posting and reliance upon list prices, the automobile industry has done a good job of managing the benchmark price against which consumers evaluate the fairness of the final price paid.
•
Manage product price trends. The single most influential reference price that consumers employ when assessing the fairness of a product price it the previous price paid for that product. As a result, it is far easier to lower prices that are too high than to raise prices that are too low.
•
Encourage favorable comparisons.
Consumers naturally compare the prices of products within and across product categories. In anticipation of this reliance upon comparisons, a firm could suggest comparisons for consumers to consider. For instance, to combat the perception of excessively high prices for its cereals, Kellogg might be welladvised to compare the daily cost of its cereals (about 30¢) to the cost of other breakfast alternatives, such as the cost of a donut or bagel (50¢ or more).
•
Avoid unfavorable comparison through product differentiation. When compact discs first appeared in the music market, they suffered from price comparisons to vinyl records. At $15, CDs were about twice the price of the alternative they were replacing. It was only after convincing consumers that such comparisons were inappropriate, due to vast improvements in sound quality and scratch resistance, that consumers accepted the higher CD prices as fair and reasonable.
Strategy #2: Actively Manage Perceptions of Cost of Goods Sold
As Scenarios #7 and #8 suggest, consumers are sensitive to a firm’s cost of goods sold. As a general rule, consumers are reluctant to pay for products they perceive to be overpriced relative to cost. 11
57
599-114
Note on Behavioral Pricing
This consideration of COGS is especially problematic for firms that operate under high fixed and low variable costs. For example, Microsoft charges hundreds of dollars for its software, yet can churn out incremental copies of that software at the cost of a floppy disk. Many consumers view this as “unfair” and, as a result, have few qualms about using pirated software. In a similar vein, some consumer label drug companies as “greedy” when those companies charge $40 or $50 for a single dose of a branded drug when comparable generics cost a fraction of the price.
Such thinking on the part of the consumer fails to appreciate the fully-loaded cost of product delivery. In the case of Microsoft, the incremental cost of producing another copy of Windows pales in comparison to the cost of developing and supporting that application. And in the case of the drug company, many years and many millions of dollars may have gone into the research and development needed to bring that drug to market. These examples highlight the need for firms to manage the consumers’ perceptions of cost of goods sold. This can be accomplished in several ways.
•
Focus attention of fully-loaded cost of goods sold.
For firms in high fixed cost/low variable cost industries, it is difficult to justify high product prices based on the incremental cost of production. Rather, the firm can focus consumer attention on the fully-loaded cost of production. For instance, if drug companies and software developers effectively communicated the high cost of product development and product support to consumers, they may successfully combat impressions of pricing “unfairly.”
•
Bundle products to obscure cost of goods sold. Some firms sell products where costs are readily apparent to the consumer (e.g., personal computer retailers). Other firms choose to bundle those very same products with additional goods or services so as to obscure the true cost of goods sold. For example, rather than sell computer components, value-added resellers sell turn-key systems. And rather than sell tickets to sporting events, travel agents sell vacation packages that include airfare, hotel and game tickets.
By bundling products, a firm can make its costs less transparent to consumers.
•
Focus attention of consumer value.
A final means by which firms can look to minimize the impact of cost of goods sold is by focusing attention on the consumer’s incentive to purchase. In the end, a consumer benefits whenever perceived value is greater than product price, regardless of the firm’s cost of goods sold. As such, through effective product positioning and communication of value, a firm can minimize the impact of COGS by maximizing attention to the net benefit of purchase to the consumer.
Summary
This note began with the statement, “It is important for a firm to get its pricing ‘right.’” Yet,
“getting prices right” is a complex process that few firms seem well-equipped to manage. While firms have a relatively good grasp of the readily available economic inputs to the pricing decision
(e.g., cost of goods sold, price of substitutes), most lack the behavioral inputs needed to fully understand and anticipate a consumers response to a pricing change. In particular, few firms anticipate the behavioral implications of transaction “fairness,” as perceived by the consumer. This note does not attempt to offer a comprehensive behavioral perspective to pricing. Rather, it is meant to raise awareness that behavioral factors exist and that these factors are important if a firm expects to
“get its pricing right.”
12
58
UVA-M-0675
A PRACTICAL GUIDE TO CONJOINT ANALYSIS
Introduction
Conjoint analysis is a marketing research technique designed to help managers determine the preferences of customers and potential customers. In particular, it seeks to determine how consumers value the different attributes that make up a product and the tradeoffs they are willing to make among the different attributes or features that comprise the product. As such, conjoint analysis is best suited for products that have very tangible attributes that can be easily described or quantified.
While the history of conjoint analysis can be traced to early work in mathematical psychology,1 its popularity has grown tremendously over the last few years as access to easy-touse software has allowed its widespread implementation. There have been probably hundreds of applications of conjoint analysis in industrial settings.2 Some of the more important questions modern conjoint analysis is used to analyze are the following:
1. Predicting the market share of a proposed new product, given the current offerings of competitors. 2. Predicting the impact of a new competitive product on the market share of any given product in the marketplace.
3. Determining consumers’ willingness-to-pay for a proposed new product
4. Quantifying the trade-offs customers or potential customers are willing to make among the various attributes or features that are under consideration in the new product design.
1
R.D. Luce and J. W. Tukey, “Simultaneous Conjoint Measurement: A New Type of Fundamental
Measurement,” Journal of Mathematical Psychology 1 (February 1964): 1–27.
2
P.E. Green, A.M. Kreiger, and Y. Wind, “Thirty Years of Conjoint Analysis: Reflections and Prospects,”
Interfaces 31 (May–June 2001): S56–S73.
This technical note was written by Associate Professor Ronald T. Wilcox. Copyright 2003 by the University of
Virginia Darden School Foundation, Charlottesville, VA. All rights reserved. To order copies, send an e-mail to sales@dardenpublishing.com. No part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the permission of the Darden School Foundation. Rev. 11/03.
59
-2-
UVA-M-0675
The Anatomy of a Conjoint Analysis
Literally, conjoint analysis means an analysis of features considered jointly. The idea is that, while it is difficult for consumers to tell us directly how much each feature of a product is worth to them, we can infer the value of an individual feature of a product by experimentally manipulating the features of a product and observing consumers’ ratings for that product or choices among competing products.
To fix your intuition here, consider the simple example of a sports car. It would be difficult for the average consumer to tell a market researcher exactly how much more valuable a car with 240 horsepower is to them relative to one with 220. It is possible that a consumer might be able to come up with some dollar value, but that value may not really reflect the way they would make choices if faced with a real marketplace situation. Instead, marketers have found that it is much more accurate to present individuals from the target market with a series of cars, described not only by their horsepower but by other attributes as well (color, price, standard/automatic transmission, etc.) and then ask them to rate each of the cars on a numerical scale. Alternatively, the researcher presents several competing cars with different attributes and asks the consumer to choose one. By repeatedly asking the potential customers to rate the cars or chose a car from a competing set, the researcher can infer the value of each individual attribute.
This is the essence of a conjoint analysis; replacing the relatively inaccurate method of asking about each attribute in isolation with a model that allows us to infer the attributes’ values from a series of ratings or choices.
The Experimental Design
A conjoint analysis begins with an experimental design. This design includes all attributes and the values of the attributes that will be tested. Conjoint analysis distinguishes between attributes and what are generally called “levels.” An attribute is self-explanatory. It could be price, color, horsepower, material used for upholstery, or presence of a sunroof, while a level is the specific value or realization of the attribute. For example, the attribute “color” may have the levels “red,” “blue,” and “yellow,” while the attribute “presence of a sunroof” will have levels “yes” and “no.” Before a researcher begins to collect data, it is important that all the levels of each attribute to be tested are written down. Commercially available software packages require that the user provide these as input.
60
UVA-M-0675
-3-
Continuing with our car example, an experimental design might look like the information presented in Table 1:
Table 1. Example of experimental design.
Levels
Price
$23,000
$25,000
$27,000
$29,000
Brand
Toyota
Volkswagen
Saturn
Kia
Horsepower
220 HP
250 HP
280 HP
Upholstery
Cloth
Leather
Sunroof
Yes
No
This is a very simple design that contains a total of 15 attribute levels. Real designs often contain more attributes and levels than are presented here.
When constructing an experimental design, it is important to keep in mind:
1. The more tangible and understandable the levels of each attribute are to the respondents, the more valid the results of the research will be. For example, attribute levels such as “really roomy” are vague, meaning different things to different people and should be avoided.
2. The greater the number of attribute levels to be tested, the more data that will be needed to achieve the same degree of output accuracy.
3. For quantitative variables (price and horsepower in this example), the greater the distance between any two consecutive levels, the harder it will be to get a good idea of how a consumer might evaluate something in between the two (i.e., $24,000).
Data Collection
Collecting data for a conjoint analysis has been made relatively simple by the advent of dedicated off-the-shelf software. The exact nature of the data collected will be dictated by the type of conjoint analysis that is used. An exhaustive discussion of the benefits and drawbacks of each of the many different types of conjoint analysis now in use is beyond the scope of this technical note. However, those interested are encouraged to read Orme for a good discussion of this topic.3
The state-of-the-art in conjoint data collection involves using personal computers or a
Web-based version of the software to guide respondents through an interactive conjoint survey.
The software creates the hypothetical product profiles using the experimental design provided by the researcher and estimates the attribute-level utilities from participant ratings or choices.
3
B. Orme, “Which Conjoint Method Should I Use?” Sawtooth Software Technical Paper (2003). Currently, an
Acrobat-readable copy of this paper is available at http://www.sawtoothsoftware.com/techabs.shtml#which.
61
-4-
UVA-M-0675
Interpreting Conjoint Results
Understanding the basic output
The basic results of a conjoint analysis are the estimated attribute level utilities. Keeping with the example in Table 1, conjoint output might look like Table 2:
Table 2. Conjoint analysis output.
Attribute
Price
Brand
Horsepower
Upholstery
Sunroof
Level
$23,000
$25,000
$27,000
$29,000
Toyota
Volkswagen
Saturn
Kia
220 HP
250 HP
280 HP
Cloth
Leather
Yes
No
Utility (Part-worth)
2.10
1.15
1.56
1.69
0.75
0.65
0.13
1.27
2.24
1.06
1.18
1.60
1.60
0.68
0.68
t-value
14.00
7.67
10.40
11.27
5.00
4.33
0.87
8.47
14.93
7.07
7.87
10.67
10.67
4.53
4.53
The estimated utilities or part-worths correspond to average consumer preferences for the level of any given attribute. Within a given attribute, the estimated utilities are generally scaled in such a way that they add up to zero. So a negative number does not mean that a given level has “negative utility”; it just means that this level is on average less preferred than a level with an estimated utility that is positive.
Conjoint analysis output is also often accompanied by t-values, a standard metric for evaluating statistical significance. Because of the way conjoint utilities are scaled, the standard interpretation of t-values can yield misleading results. For example, the level “Saturn” of the attribute “Brand” has a t-value of 0.87. In general, a t-value of this magnitude would fail a test of statistical significance. However, this t-value is generated because within the attribute “Brand” the level “Saturn” has neither a very high nor very low relative preference. It is basically in the middle in terms of overall preference. Because of the scaling, levels that have more moderate levels of preference within a given attribute are likely to have estimated utilities close to zero, which will tend to produce very low t-values (recall that the t-test is measuring the probability that the true value of a parameter is not different from zero).
62
-5-
UVA-M-0675
A better way to think about statistical significance in this context is to examine the tvalues of the levels with the highest and lowest preference within a given attribute. An applicable common practice would be if the sum of the absolute value of these two statistics is greater than three, then that given attribute is significant in the overall choice process of consumers. At a practical level, it is rare that an attribute will not be significant, and, if you find one that is, it means that it probably should not have been included in the experimental design in the first place, because respondents are not considering that attribute’s information when they make choices. Conjoint Analysis Applications
As mentioned previously, there are many different possible applications of conjoint analysis. We will focus on three very common applications: trade-off analysis, predicting market share, and determining overall attribute importances.
Trade-off analysis
The utility of any given product that we might consider can be easily computed by simply summing the utilities of its attribute levels. For example, a Toyota with 280 horsepower, leather interior, no sunroof, and a price of $23,000 has a utility of 0.75 + 1.18 + 1.60 0.68 + 2.10 =
4.95. If the car with same basic specifications were a Volkswagen, the overall utility would drop to 0.65 + 1.18 + 1.60 0.68 + 2.10 = 4.85, a drop of 0.10. This drop can be seen directly by noticing that the difference between the utility for the brand Toyota (0.75) and Volkswagen
(0.65) is 0.10 and because nothing else in the profile of the car has changed this will be the exact utility difference between two cars that are the same except for this brand difference.
A natural consequence of the above observation is that we use the utilities to analyze what average consumers would be willing to give up on one particular attribute to gain improvements in another. For example, how much money would they be willing to give up
(price) if a sunroof was added to the vehicle? We will now look directly at this issue of the hypothetical car detailed in the previous paragraph. Adding a sunroof to the (Toyota) car would yield an overall utility of 0.75 + 1.18 + 1.60 + 0.68 + 2.10 = 6.31. This represents an increase in utility of 6.31 – 4.95 = 1.36 over the identical car without a sunroof.
The information above directly implies that we can reduce the utility of price by 1.36, and average consumers would be just as happy as before the sunroof was installed. To find out how much the price can be raised, we must convert the change in utility with a change in price.
We do this by first noting how much the original car costs ($23,000) and the utility associated with that figure, 2.10. We know that we can reduce the price utility by 1.36. This is equivalent to saying that we can reduce the price utility to 2.10 – 1.36 = 0.74. By referring to Table 2, we can immediately see that this implies a price between $25,000 and $27,000 because 1.56 < 0.74 <
1.15. In fact, if we assume a linear relationship between price and utility in the range between
63
UVA-M-0675
-6-
$25,000 and $27,000, we can solve for the exact price by performing a linear interpolation within this range.4 Specifically, the interpolation yields:
$25,000
1.15 0.74
$2,000
1.15 ( 1.56)
Utility spread between the two tested price points (25K and 27K).
$25,302.58
Utility spread between 25K and the target utility.
This implies that, if the sunroof is added, the price of the vehicle could be raised from
$23,000 to about $25,300, and the average consumer’s attitude would be one of indifference between the two vehicles. Qualitatively, it shows that the value of a sunroof to consumers is very substantial. This same kind of analysis can be performed for other attributes. We could ask how much additional horsepower we would need to add if the interior was changed from leather to cloth.
This particular question does present a problem, however. Because the current vehicle under consideration has 280 horsepower, and that is the maximum amount of horsepower tested by the conjoint analysis, it will be impossible to determine how much consumers will value additional horsepower. This leads to an important consideration when constructing the experimental design.
That is, if the output is to be used for trade-off analysis, it is important that the range of the levels tested within each attribute span the entire range of that attribute before management would ever consider it as a realistic design alternative. If the experimental design takes this into account, we can perform trade-off analysis between any two attributes in the design.
Market share forecasting
Another common application is forecasting market share. In order to use conjoint output for this kind of prediction, two conditions must be satisfied:
1. The company must know the other products, besides their own offering, that a consumer is likely to consider when making a selection in the category.
2. Each of these competitive products’ important features must be included in the experimental design. In other words, you must be able to calculate the utility of not only your own product offering but that of the competitive products as well.
4
This is a common way to approximate the relationship between the value of the attribute and its utility for attribute values that were not directly tested by the conjoint analysis. The closer the tested levels are to each other the more accurate this approximation. Also notice that this interpolation can only be performed for quantitative attributes such as price. Interpolating between qualitative attributes, like brand, is nonsensical.
64
UVA-M-0675
-7-
Market share prediction relies on the use of a multinomial logit model.5 The basic form of the logit model is: eU i
Sharei
n j 1
Uj
e
where:
Ui is the estimated utility of product i
Uj is the estimated utility of product j n is the total number of products in the competitive set, including product i.
To make things clear, consider the following example. Suppose we are interested in predicting the market share of a car with the following profile: Saturn; $23,000; 220 HP; cloth interior; no sunroof. We believe that when consumers consider our car they will also consider purchasing cars that are currently on the market with the following profiles:
1. $27,000; Toyota; 250 HP; cloth interior; no sunroof
2. $29,000; Volkswagen; 280 HP; leather interior; no sunroof
3. $23,000; Kia; 220 HP; cloth interior; no sunroof
For the Saturn and its associated product profile the estimated utility is 2.10 – 0.13 – 2.24
– 1.60 – 0.68 = 2.55. Similarly, the utilities of the three competing products can be calculated:
1.
2.
3.
1.56 + 0.75 + 1.06 – 1.60 – 0.68 = 2.03
1.69 + 0.65 + 1.18 + 1.60 – 0.68 = 1.06
2.10 – 1.27 – 2.24 – 1.60 – 0.68 = 3.69
With these utilities in hand we can now directly apply the logit model to forecast market share for the Saturn. This is given by:
ShareSaturn
e e 2.55
e
2.03
2.55
e1.06
e
3.69
.025 or 2.5%
This implies that this particular Saturn vehicle will achieve a 2.5% market share within the specified competitive set. The market share of any vehicle that can be described by the experimental design and a set of competitive vehicles, also described by the experimental design, can be found in a similar manner.
5
A good marketing reference to learn about the basics of the logit model is G. Lilien and A. Rangaswamy,
Marketing Engineering: Computer-Assisted Marketing Analysis and Planning (2nd Ed.), Englewood Cliffs:
Prentice-Hall (2002). Also, most econometrics textbooks will have information on logit models.
65
UVA-M-0675
-8Determining attribute importance
A researcher may also be interested in determining the importance of any individual attribute in the consumers’ decision processes. Quantifying these attribute importances using the conjoint output is straightforward and can provide both interesting and useful insights into consumer behavior.
Intuitively, the variance of the estimated utilities within a given attribute tells you something about how important the attribute is in the choice process. Take, for example, the attributes “Sunroof” and “Upholstery,” both of which have only two levels. If you understand the material up to this point, it should be reasonably clear that “Upholstery” is a more important attribute than “Sunroof.” That is because the utility difference between having a sunroof and not having a sunroof (2 × 0.68 = 1.36) is smaller than the utility difference between having leather versus cloth interior (2 × 1.60 = 3.20).
The common metric used to measure attribute importances is:
Ii
Ui n j 1
Ui
Uj
U
j
where:
Ii = the importance of any given attribute i
U = the highest utility level within a given attribute (subscripts indicate which attribute)
U = the lowest utility level within a given attribute.
This equation is really quite intuitive. In order to calculate the importance of any given attribute, you just take the difference between the highest and lowest utility level of that attribute and divide this by the sum of the differences between the highest and lowest utility level for all attributes (including the one in question). The resulting number will always lie between zero and one and is generally interpreted as the percent decision weight of an attribute in the overall choice process.
It also should be clear at this point that this estimated attribute importance depends critically on your experimental design. In particular, if you increase the distance between the most extreme levels of any given attribute you will almost certainly increase the overall attribute importance. For example, if the tested price range was $21K $31K instead of $23K – $29K
(Table 1), this is very likely to increase the estimate attribute importance of price.
Let’s now consider a concrete example using the attribute “Horsepower.” The importance of this attribute is calculated as:
66
-9-
I Horsepower
UVA-M-0675
1.18 2.24
= .25
(2.10 1.69) (0.75 1.27) (1.18 2.24) (1.60 1.60) (0.68 0.68)
In our example, 25% of the overall decision weight is assigned to horsepower. The reader may verify through analogous calculations that the decision weight for “Price” is about 27%,
“Brand” about 15%, “Sunroof” about 10%, and “Upholstery” about 23%. The numbers provide a very intuitive metric for thinking about the importance of each attribute in the decision process.
Final Thoughts
Conjoint analysis has a broad array of possible applications. Many of these applications are variants of the three very common applications presented here. The increasingly widespread availability of conjoint analysis software, both PC and Web-enabled, points to its continued growth as a marketing decision aid.
This note has presented what is generally known as “aggregate-level” conjoint analysis.
That is, all of the respondents are pooled into one group, and a single set of attribute level utilities are estimated from the ratings or choices provided by the people in this group. Recent advancements in conjoint analysis have enabled researchers to estimate different utilities for different groups of respondents and even, in some cases, for individual respondents. While the mathematics necessary for this procedure is sometimes quite complex, it is now possible to estimate the attribute level utilities and to compute trade-off analyses for each individual respondent. This has some significant advantages over aggregate level analysis particularly when considering marketing segmentation issues. Either way, the data collection and the basic interpretation of the output remains the same. While there is currently no textbook that can provide the reader with answers to all of the questions that might arise when applying this technique in a business setting, there is, as of this writing, a very good and surprisingly comprehensive collection of technical papers located on the site of a company that markets conjoint analysis software (http://www.sawtoothsoftware.com/techpap.shtml). These provide answers to many of the practical implementation questions a user may face.
67
68
UVA-M-0674
FIDELITY INCORPORATED:
PRICING THE FIDELITY BLUE CHIP GROWTH FUND
John McDowell walked off of the 18th green at Andover Country Club and handed his putter to his caddie in obvious disgust. He had double-bogeyed the 18th and, in doing so, shot an
84, his worst round of golf in the last several months.
McDowell knew what the problem was. It was not his stance, his shoulder position, or even his often-beguiling driver. It was his lack of concentration brought on by the pricing decision facing him, a decision upon which he knew millions of dollars were at stake.
As the fund manager of Fidelity’s Blue Chip Growth Fund (BCGF), McDowell not only oversaw the investment decisions of the fund but had considerable input into its marketing plan as well. Senior management at Fidelity had asked McDowell to consider how to generate more profit from the fund. While the fund had grown significantly over the last few years, management believed that they were not charging customers enough for the services provided.
Fidelity had recently spent a great deal of money updating and enhancing their on-line order handling and customer service capabilities and believed it was now appropriate to ask their customers to pay something for these increased services.
BCGF currently had fees that were significantly below the industry averages for actively managed domestic equity mutual funds. Investors in this fund paid no front- or back-end load and annual expenses of 77 basis points (.77 percent).1 Total expenses attributed to the fund and expenses for fund management and marketing came to about 45 basis points. As of May 2003,
BCGF held $17.9 billion in assets.
Most BCGF investors viewed management of their investments as a reasonably long-run decision. While the average holding time for an investment in the fund fluctuated with economic and market conditions, management believed the average holding time to be around seven years.
This number was typically larger for investors whose money was tied to a 401(k) or other tax1
A front-end load is a one-time charge, generally expressed as a percent of the assets invested (i.e., 2 percent).
A back-end load is analogous to a front-end load except the fee is paid when fund shares are redeemed. Annual expenses, also expressed as a percent of assets under management, are deducted from the fund on an ongoing basis.
This case was prepared by Ronald T. Wilcox, associate professor of Business Administration, Darden Graduate
School of Business Administration, University of Virginia. It was written as a basis for class discussion rather than to illustrate effective or ineffective handling of an administrative situation. Copyright 2003 by the University of
Virginia Darden School Foundation, Charlottesville, VA. All rights reserved. To order copies, send an e-mail to sales@dardenpublishing.com. No part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the permission of the Darden School Foundation.
69
-2-
UVA-M-0674
defined contribution pension plan and typically less for those holding the funds in a regular, nontax-advantaged, brokerage account. In the previous year about 15 percent of the mutual fund shares held by investors at the beginning of the year were sold, and the money moved elsewhere.
The fund also received new monies in the amount of about 20 percent of the value of the fund.
Both of these figures represent dollars-in-the-door and dollars-out-of-the-door based on customers’ decisions and are not related to any investment gains or losses experienced by the fund. So, the fund experienced a net cash inflow of 5 percent during the previous calendar year.
While these types of inflows and outflows were difficult to predict, McDowell believed a similar pattern might hold in the coming year. Table 3 provides some information on the growth of mutual fund ownership among U.S. households.
Edward Johnson, CEO of Fidelity Investments, in a memo to McDowell earlier in the week, had asked him to explore ways to increase fund profitability by 30 percent. It was this task that had McDowell concerned. He didn’t know whether he should try to generate the additional profit by introducing a front-end load to the fund or by increasing the annual fees.
Back-end loads were probably not a feasible alternative given the history and pricing philosophy of Fidelity. And, although any front-end load was technically possible, industry norms dictated that loads were set in increments of 50 basis points. So, a load of 1.5 percent was fine in terms of industry norms, but a load of 0.75 percent was not.
McDowell was concerned that potential customers might view the higher fees as a deterrent to investing in BCGF, and worried that even current customers would leave the fund.
He knew he needed to find a way to generate the additional profit in a way that would lead to the least dissatisfaction among current and potential future investors.
McDowell’s assistant, Maria Ramos, had pointed out a recently published study that might help him make the new pricing decision. The study used the marketing research technique, conjoint analysis, to examine investors’ preference for stock mutual funds. This particular conjoint study had estimated the utilities or preferences of investors for different levels of front-end loads and annual expenses among other fund attributes. Ramos pointed out that the results of this study might shed light on how to increase profitability without seriously harming present or future investor satisfaction with the fund.
The Conjoint Study
A detailed description of the conjoint analysis of mutual fund choice was published in a technical academic journal.2 Ramos distilled what she believed to be the important facts for the decision at hand and brought them to McDowell. These facts included the important features of the experimental design, given in Table 1 and the results of the conjoint estimation, given in
Table 2. The headings of the columns in Table 1 indicate the name of the tested attribute and the values below the heading the levels of each of the tested attributes. Table 2 provides the estimated utilities or part-worths for each level of each attribute.
2
R.T. Wilcox, “Bargain Hunting or Star Gazing? Investors’ Preferences for Stock Mutual Funds,” Journal of
Business 76 (2003).
70
-3-
UVA-M-0674
Table 1: Experimental Design
Front-End Load
Annual Fee
10-Year Past Return
Beta3
No Load
0%
5%
0.7
1%
0.5%
10%
0.9
2%
1.0%
15%
1.1
3%
1.5%
20%
1.3
4%
2.0%
25%
1.5
5%
2.5%
________________________________________________________________________
Company
Fidelity
Vanguard
T. Rowe Price
Dreyfus
Pecunia
The study appeared to be conducted using a representative sample of mutual fund investors, but not necessarily customers of Fidelity BCGF or any other Fidelity mutual fund.
Ramos reported to McDowell that the data was collected using a specific type of conjoint called
Choice-based Conjoint analysis, a type that seemed appropriate for this application.4 Among the other results of the study, the author had also concluded that investors appeared not to take into account their investment time horizon when evaluating the fees of a given fund.
In addition to the question of price, McDowell also believed that Fidelity’s brand name was very strong and, based on this brand name, he could probably charge a premium to his chief competitor, Vanguard. Vanguard had been aggressively marketing their PRIMECAP fund, a fund with a very similar asset base, to BCGF. PRIMECAP charged no front-end load and had an annual fee of 49 basis points. McDowell believed that Fidelity’s superior customer service, investment in advertising, and enhanced Web interface sufficiently increased investors’ perception of the Fidelity brand to insulate BCGF from any serious price competition from
Vanguard.
The Pricing Decision
The stakes for this particular pricing decision were huge. McDowell, scribbling some quick calculations on a napkin, realized that every additional basis point they were able to capture in profit would increase fund profitability by almost $2 million. He struggled with how to use the information from the conjoint analysis to help his pricing decision. He wondered whether a study published in some academic journal he had never heard of, using a sample of investors who were not necessarily his customers, was useful at all in this situation. He felt confident that the Fidelity brand would keep him from losing investors even in the face of the price increase. Ultimately, he told himself, the increased profit target of 30 percent is not my call; my job is to figure out how to get this done.
3
Beta is a measure of the covariance of the return of a fund relative to an index, in this case the S&P 500. It is often used as a measure of the volatility of a fund. For example, a beta of 1.3 indicates that the fund is 1.3 times more volatile than the S&P 500.
4
B. Orme, “Which Conjoint Method Should I Use?” Sawtooth Software Technical Papers (2003).
71
UVA-M-0674
-4Table 2: Results of Conjoint Analysis
Utility Estimate
St. Dev. of Estimate
Fidelity
Vanguard
T. Rowe Price
Dreyfus
Pecunia
0.08
0.01
0.08
0.00
-0.17
0.09
0.09
0.09
0.09
0.09
No Load
1%
2%
3%
4%
5%
0.88
0.48
0.05
-0.25
-0.49
-0.67
0.10
0.10
0.11
0.11
0.13
0.13
0%
0.5%
1.0%
1.5%
2.0%
2.5%
0.99
0.50
0.05
-0.07
-0.72
-0.74
0.10
0.10
0.11
0.11
0.13
0.13
5%
10%
15%
20%
25%
-1.23
-0.70
-0.06
0.76
1.23
0.10
0.10
0.10
0.10
0.10
0.7
0.9
1.1
1.3
1.5
0.05
0.06
0.13
-0.10
-0.14
0.10
0.10
0.10
0.10
0.10
Company
Load
Annual Expenses
10-Year Return
Beta
72
UVA-M-0674
-5Table 3: The Growth in Mutual Fund Ownership
U.S. Households Ow ning Mutual Funds, 1980-2001
52.0%
49.0%
44.0%
37.2%
30.7%
24.4%
25.0%
27.0%
20.0%
11.0%
11.9%
5.7%
1980 1982
1980
4.6
1982
9.0
1984
1984
10.2
1986 1988
1986
17.3
1988
22.2
1990 1992
1990
23.4
1992
25.8
1994 1996
1994
30.2
1998
1996
36.8
2000 2001
1998
44.4
2000
51.7
Millions of U.S. Households
Source: Investment Company Institute, “Fundamentals: U.S. Household Ownership on Mutual Funds in 2001.”
73
2001
54.8
74
UVA-M-0773
Rev. Feb. 22, 2010
PORTLAND TRAIL BLAZERS
“We have a new group sales operation in place and we’re looking for dramatic results there.”
—Steve Patterson, President
Portland Trail Blazers
Less than a month after the 2005 NBA All-Star break, the Portland Trail Blazers was a team in upheaval. On the court, they had just fired their coach of the past four seasons and were
22–36, in danger of one of the worst seasons in franchise history. Off the court, the Blazers organization was facing considerable challenges as well. The team’s home arena, the Rose
Garden, had filed for Chapter 11 bankruptcy and was being run by the building’s creditors.
The arena, a virtual lock to sell out just three seasons ago, had seen attendance numbers fall more than 15% since the 2003 season (Figure 1). During the same time, the organization had only been successful in renewing 9 of the 46 luxury-suite contracts that came due in 2005, and
42 of the 70 luxury suites sat empty during the season.1 Television interest also declined, with a
Portland-area Nielsen share of just 5% when the Blazers played the Minneapolis Timberwolves
(weather coverage generally received up to 20%).2
1
2
Todd Murphy, Todd. “Have Arena, Need People.” Portlandtribune.com, August 10, 2004.
Pete Schulberg, “Blazers Start Losing with Viewers, Too,” Portland Tribune, January 21, 2005.
This case was prepared by Professor Ron Wilcox. It was written as a basis for class discussion rather than to illustrate effective or ineffective handling of an administrative situation. Copyright ¤ 2009 by the University of
Virginia Darden School Foundation, Charlottesville, VA. All rights reserved. To order copies, send an e-mail to sales@dardenbusinesspublishing.com. No part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the permission of the Darden School Foundation. Rev. 2/10. ¸
75
-2-
UVA-M-0773
Figure 1. Average home attendance, 1996–2007.
Attendance
21,000
20,000
19,000
15%
18,000
17,000
16,000
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Season
Source: Portland Trail Blazers.
A similar story was occurring in the sale of the team’s “club seats”—special seats for
Blazers games that were sold on a multiyear contract and came with club perks. Of 1,800 club seats in 2005, 700 remained available, most of them because subscribers had dropped their contracts during the previous season.
The Portland Sports Market
The Trail Blazers had a monopoly on the professional sports market in Portland, Oregon.
Without a dominant university sports program affiliated with the city, the team competed only with minor league baseball and hockey for its share of the city’s sporting dollars. At just under two million people in the metro area, Portland was the fourth-smallest market in the NBA.
Each year the Blazers tracked Portland-area residents’ general perception of the team (see
Figure 2). The historically strong relationship between the team and the city had soured over the past few seasons, with the percentage of people perceiving the team negatively having increased tenfold since 2000. Fan support had dwindled due to a number of widely publicized player transgressions including marijuana use, fighting with teammates, and an incident involving animal cruelty.3
3
Andy Giegerich, “Beleaguered Blazers Play by the Numbers,” Portland Business Journal, October 29, 2004.
76
-3-
UVA-M-0773
Figure 2. Portland-area residents’ general perception of the Portland Trail Blazers.
100%
92%
Favorable
Unfavorable
80%
71%
60%
40%
Jun-01
Dec-01
60% 63%
56%
53%
51%
42%
47%
49%
29%
24%
20%
5%
0%
Jun-00 Dec-00
68%
Jun-02
Dec-02
Jun-03
54%
46%
40% 37%
Dec-03
Jun-04
57%
43%
Dec-04
Source: Portland Trail Blazers.
Multigame Ticket Packages
One of the more successful Blazers promotions during the past few seasons had been multigame ticket packages. This program allowed fans to purchase tickets for a number of games at once, and usually included at least one marquee opponent, a game for which individual tickets were difficult to find. Trail Blazers’ management saw this program as an effective tool to:
1. Increase ticket sales for less popular games (typically bundled in the package with tickets for hard-to-find games)
2. Increase overall ticket sales because the multigame packages acted as an effective incentive for those who planned on attending only one or two games during the season to increase the number of games they attended
3. Develop more of an ongoing relationship with fans who could potentially become future season ticket holders
Despite the program’s relative success, management wanted to explore all potential packages, and better understand which options were most popular with fans. The program’s goal was to offer a multigame ticket package that had a high appeal to fans, while being profitable to the team and not undermining current pricing policies.
77
-4-
UVA-M-0773
Designing the Research Study
The Trail Blazers management team hired Acuity Market Research, a Portland-based research firm, to help design their multigame package study. Together, they determined there were six aspects of the multigame ticket packages that drove a customer’s decision to purchase:
1. The team the Blazers were playing
2. The day of the week the game was being played
3. The number of games included in the package
4. The location of the seats
5. The price (per seat) of the package
6. The promotional item included in the package
The project team designed a study utilizing conjoint analysis to ascertain the importance of the individual attributes, as well as the likely response of the market to specific multigame ticket packages. Some things were givens: (1) there was a high number of teams in the NBA (30 including the Blazers) and (2) the dates of the games included in the package could not be changed, those attributes would not be included as part of the conjoint products. Instead, questions pertaining to favorite teams and days to watch a game were asked individually, after the conjoint portion of the survey.
An e-mail went out from the Blazers’ director of database and Internet marketing to 960 fans who had purchased multiple-game ticket packages or season tickets in the past, but were not current season ticket holders. The project team decided it was more important to get feedback from people who had expressed some level of commitment to purchasing Blazers tickets in the past than general fans of the team, and knew it had current e-mail addresses for this group.
Although new fans were always purchasing multigame or season ticket packages, the Blazers’ management believed past purchasers were likely the best prospects for new multigame packages. The initial e-mail explained the purpose of the study and asked fans to participate. One week later, a reminder e-mail was sent in hope of increasing the overall response rate of the study. Both e-mails contained a link to an online conjoint-based survey which included 20 different conjoint choice tasks, (an example is included in the Appendix) Blazers-specific questions, and a battery of demographic questions. Most respondents took between 10 and 15 minutes to complete the survey.
Participants were given a chance to win free tickets to Blazers games, as well as autographed Blazers items such as jerseys, basketballs, and posters for taking part in the survey.
78
-5-
UVA-M-0773
Study Findings
The e-mail solicitations received a total of 204 valid responses (a 21% response rate).
Summary statistics regarding demographics and past Blazers-game attendance are located in the
Appendix.
Acuity began their analysis of the multigame packages by computing the attribute level utility scores to help better understand the stand-alone preference of each of the individual attribute levels. The utility score data are shown in Table 1.
Table 1. Utility score data.
Utility
0.03257
0.24383
–0.2764
Utility
0.65646
0.22011
0.126
–1.00257
Utility
–0.73169
–0.43716
0.15736
1.01148
Utility
0.12511
0.17428
0.00158
–0.31786
0.01689
Number of Games
3-game create-your-own pack, including 1 elite team and 2 very good teams
6-game create-your-own pack, including 2 elite teams and 4 very good teams
10 game create-your-own pack, including any combination of teams.
Ticket Price
$15 per seat per game
$25 per seat per game
$35 per seat per game
$60 per seat per game
Ticket Location
300 level, behind the baskets
300 level, on the corners
300 level, midcourt
200 level, midcourt
Promotional Item
Priority for home playoff tickets
Hot dog and soda with each ticket
Trail Blazer apparel (hat, jersey, etc.)
Trail Blazer collectible (bobblehead, etc.)
$20 gift certificate for popular local restaurant
Source: Portland Trail Blazers.
While the conjoint study allowed all the attributes and levels to be randomly assigned, in reality the Blazers were unwilling to allow certain price and seating combinations no matter how well received they were, due to the cost structure of the arena. They disallowed: x 200-level seats for less than $60
x
300-level, midcourt seats for less than $25
79
-6-
UVA-M-0773
Cost of Multigame Packages
While the fan preference was extremely important to Blazers’ management, any multigame packages the group designed had to be financially attractive and align with the organization’s strategic goals. Each of the multigame ticket package attributes had costs and strategic implications associated with them:
Number of Games: The Blazers prefer the six-game package because it offered the capability of pairing the most popular teams with games that were more difficult to sell tickets to
(weekday games, less competitive teams, etc.). Their next preference would be a 10-game package, because it allowed the team to efficiently sell a large number of the remaining games to a single fan.
Seat Location: Although nearly all the Blazers’ stadium costs were fixed expenses, the organization still applied a cost to each of the seat locations in the stadium. This cost structure had to be met, at a minimum, for any tickets that were sold and differed based on seat location.
Minimum seat pricing is shown in Table 2.
Table 2. Fixed costs based on seat location.
Seat Location
Fixed Cost
300 level, behind the baskets
$10.00
300 level, on the corners
$12.00
300 level, midcourt
$18.00
200 level, midcourt
$40.00
Source: Portland Trail Blazers.
Promotional Items: A direct cost is associated with each of the promotional items the
Blazers may offer fans. For example, if the Blazers were to offer a hot dog and soda with each ticket, they would have to pay the Rose Garden’s vendor services a negotiated price of $3.25 per package. The $20 gift certificate to a popular restaurant is purchased for a negotiated price of
$10. The restaurateur deeply discounts the gift certificates in exchange for the marketing exposure. The only promotional item without a direct cost was offering priority for home playoff tickets, given that the tickets are still sold at full retail price and multigame ticket holders just receive priority in purchasing available tickets. Table 3 presents the unit cost of each of the potential promotional items.
80
-7-
UVA-M-0773
Table 3. Cost of promotional items.
Promotional Item
Cost
Priority for playoff tickets
$0.00
Hot dog and soda with each ticket
$3.25
Trail Blazers apparel (hat, jersey, etc.)
$12.00
Trail Blazers collectible (bobble head, etc.) $6.75
$20 gift certificate to a popular restaurant
$10.00
Source: Portland Trail Blazers.
Utilizing the conjoint information, in addition to the other data available from the survey, the Blazers’ management team felt prepared to design the multigame package they believed the fans would most prefer. What attributes were most important to the fans? What should the
Blazers’ multigame package include? Should there be more than one? How profitable would each of the packages be?
81
-8Appendix
PORTLAND TRAIL BLAZERS
Example: Online Conjoint Survey
Source: Portland Trail Blazers.
82
UVA-M-0773
-9-
UVA-M-0773
Appendix (continued)
Study Demographics
People under the age of 18 in the home
Gender
Female,
21.9%
57.5%
Zero
Male,
78.1%
22.5%
One
15.5%
Two
3.5%
Three
1.0%
Four or more
0
0.1
0.2
0.3
0.4
0.5
10.4%
18-24
65.2%
Married
25-34
27.8%
26.9%
30.8%
35-44
20.4%
45-54
6.1%
Divorced
55-64
1.0%
Widowed
0
6.5%
5.0%
65 and over
0.1
0.2
0.3
0.4
0.5
0.6
0.0%
0.7
5.0%
Under $20,000
3.5%
Some College (including vocational and technical school)
26.4%
0.2
0.3
6.3%
$30,000 to $39,999
6.3%
25.0%
30.0%
35.0%
7.9%
9.4%
$60,000 to $69,999
11.5%
25.1%
$75,000 to $99,999
20.4%
0.1
20.0%
2.6%
$50,000 to $59,999
49.8%
0
15.0%
$20,000 to $29,999
$40,000 to $49,999
College Graduate
Advanced Degree
10.0%
Annual Household Income
Educational Level of Respondents
High School Graduate
0.7
Age of Respondents
Marital Status of Respondents
Single
0.6
30.9%
$100,000 or more
0.4
0.5
0.6
0
Source: Portland Trail Blazers.
83
0.05
0.1
0.15
0.2
0.25
0.3
0.35
-10-
UVA-M-0773
Appendix (continued)
Game Attendance Behavior
Have purchased multi-game tickets
Have purchased single-game tickets
No, 23.2%
No, 41.9%
Yes, 58.1%
Yes, 76.8%
Have purchased season tickets
Have purchased other tickets
Yes, 4.0%
Yes, 15.2%
No, 84.8%
No, 96.0%
Number of Trail Blazer home games attended in past two seasons
None
2.7%
1-2
3-5
5.9%
20.0%
6-10
11-15
13.5%
16-20
21-30
31-40
More than 40
0.0%
23.2%
20.5%
3.8%
8.6%
1.6%
5.0%
10.0%
Source: Portland Trail Blazers.
84
15.0%
20.0%
25.0%
85
Source: Portland Trail Blazers.
Rose Garden Seating Chart
Appendix (continued)
-11-
UVA-M-0773
86
9-503-019
AUGUST 7, 2002
ELIE OFEK
Customer Profitability and Lifetime Value
The digital revolution has given rise to a host of technologies that are transforming marketing practices. Powerful databases and electronic data networks are allowing companies to collect concise information about customers and their buying patterns more effectively and efficiently than ever before. The Internet, in particular, has increased the ability of firms to track the behavior of individual consumers as they visit numerous Web pages.
While these capabilities challenge companies to manage vast amounts of information, they offer marketers exciting new opportunities to dynamically manage their customer bases. Beyond the traditional focus on mass advertising, firms can now more accurately recognize and monitor individual customers to whom tailored communications and offers can be made. Marketing activity is becoming more interactive and engaging. For many companies, particularly large or multinational corporations, this was altogether infeasible or prohibitively costly in the past.
In order to reap the benefits of detailed customer knowledge, firms need to systematically estimate the profitability associated with its use. The ultimate goal is to develop highly committed customers who not only make repeat purchases and generate continual revenue streams, but also require minimal maintenance along the way. It is entirely possible that while some customers do not bring profits with their initial purchases, the margins from their future expected transactions paint a different picture. As a result, firms need to track initial acquisition costs and compare them to the profits to be generated over the customer’s expected relationship with the company. The above activities allow marketers to decide which customers to go after, how to change the promotional mix as a function of past and recent transactions and, if necessary, which specific customers to discontinue serving. Indeed, many practitioners and scholars are expressing the view that marketing is rapidly becoming “the science and art of finding, retaining, and growing profitable customers.”1
The initial step in determining customer profitability is to have a clear sense of the relevant characteristics of customer activity. To this end, analysis of historical data can be very powerful in providing firms with information on the buying patterns of both new and existing customers (these may or may not be correlated with certain demographics). Using this information, the firm may choose to distinguish between segments of customers that exhibit similar characteristics and are hence expected to respond similarly to a given level of communication/promotion. Taking into account the responsiveness of particular customers to certain offerings, the cost of making these offerings, and the probability a customer is expected to keep generating revenue for the firm enables
1 Philip Kotler and Gary Armstrong, Principles of Marketing (Upper Saddle River, NJ: Prentice Hall, 2001).
________________________________________________________________________________________________________________
Professor Elie Ofek prepared this note with the assistance of Research Associate Kevin Morris as the basis for class discussion rather than to illustrate either effective or ineffective handling of an administrative situation.
Copyright © 2002 President and Fellows of Harvard College. To order copies or request permission to reproduce materials, call 1-800-545-7685, write Harvard Business School Publishing, Boston, MA 02163, or go to http://www.hbsp.harvard.edu. No part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the permission of Harvard Business School.
87
503-019
Customer Profitability and Lifetime Value
the calculation of customer profitability and lifetime value. These ideas are made clearer through the following example.
Example: The Catalog Retailer
Part I—Customer Acquisition Costs
A direct catalog retailer of fashion goods is contemplating whether or not to attract new customers using names purchased from a list broker or by randomly sending out catalogs. The cost of sending a catalog (which includes production and mailing) is $0.5. From experience, the company anticipates the response rate (the percentage of individuals who receive the catalog and purchase from it, hence becoming “current customers”) from a random mailing to be 1%. By analyzing the buying behavior and demographics of current customers, the retailer estimates it can rent names selectively from the broker to achieve a response rate of 4%. The list broker has set a price of $0.2 for each name. The direct retailer wishes to calculate the acquisition costs associated with each source of potential customers. To determine acquisition costs, the retailer needs to use the response rate and cost of sending a catalog to each prospect.2 A response rate of 1% (or 0.01) means that of 100 catalogs sent, only one recipient is expected to respond. Similarly, a response rate of 4% means that out of 25 catalogs sent, one recipient is expected to respond with a purchase. The cost of acquiring a customer can then be calculated using the following:
Acquisition Cost= (# of catalogs needed to get 1 customer) * (cost of sending a catalog) =
cost of sending a catalog response rate
Thus, the acquisition cost using the random mailing approach is: 0.5/.01=$50, and the cost of acquiring a customer from the rented list is: 0.7/0.04= $17.5. The retailer concludes that even though each rented name increases the effective cost of sending out each catalog by 40%, the improvement in targetability makes this a worthwhile avenue to pursue.
Additional Comments:
a) In some instances, the response rate is not given directly but can be calculated using other pieces of information. For example, direct marketers often refer to subsets of customers by the “demand per book” they generate. Demand per book is the expected revenue from one mailed catalog. If the average order size (in dollars) is known, the response rate can be calculated by dividing these two quantities, that is, response rate = (demand per book)/(average order size).
b) As any rented list of names includes a finite number of prospects that meet certain criteria, the retailer must also take into account that at some point all relevant names will be exhausted. At this point, to acquire new customers, the retailer can either move on to names expected to yield lower rates of return (but that are still profitable), or attempt to find new sources of names that were not on the previous broker’s list.
2 The marketing effort to acquire new customers from a pool of potential individuals is often called “prospecting.”
2
88
Customer Profitability and Lifetime Value
c)
503-019
The above calculation of acquisition costs can be generalized to other industry contexts. To do so, one needs to recognize the instruments used to generate a response (such as a visit by a salesperson, a call from a telemarketer, mailing of an informational brochure, free samples, or special discount coupons) and the likelihood of obtaining a response.
Part II—Customer Break-Even Analysis
Now that we have determined the cost of acquiring a customer, we can use this information to establish how many purchases or years it will take for each customer to realize profits for the firm.
In our example, the catalog retailer makes changes to the product line four times a year. Historical data has shown that one can roughly segment the market of current customers after one year into frequent buyers, who purchase twice a year with an average order size of $50, and occasional buyers, who purchase only once a year with an average order size of $80. The retention rate (percentage of customers who continue to make a purchase with the company in the next period) is 75% for frequent buyers and 50% for occasional buyers. Thus, the likelihood an individual acquired today will remain a customer by year five, or the survival rate of year five, is approximately 30% for the frequent buyers and 6% for the occasional buyers.3 Gross margins are 20% of sales, and include all expenses aside from the cost of sending out catalogs. In the first year, the retailer sends a catalog every month (with the same catalog sent in three consecutive months) to all acquired customers. Based on first year purchase patterns, frequent buyers continue to receive 12 catalogs a year, while occasional buyers receive only four. The catalog retailer is interested in knowing how many years it would take to recoup initial costs of acquisition, assuming customers were acquired using rented names.
To address the retailer’s question, it is useful to construct the following table for each of the two segments: Figure A
Customer Break-Even Analysis for Frequent and Occasional Buyers
In both tables below: In line D, expected profits in each year are obtained by multiplying the survival rate by the total margin per customer net of catalog mailing costs. For line E in Year One, the acquisition cost ($17.5) is subtracted from the sum of expected profits per customer. In subsequent years, the cumulative profits per customer of the previous year are added to the total expected profit per customer of that year.
A.
B.
C.
D.
E.
Frequent Buyers
Year One
Year Two
Margin on each purchase
Survival rate
Cost of mailing catalogs
Total expected profit per customer
Cumulative profits per customer
(net of acquisition costs)
$10
100%
.5*12=$6
2*10-6=$14
$10
75%
$6
.75*(20-6)=$10.5
$(3.5)
$7
3 For frequent buyers: .75(5-1) =.754 ≅.30, and for occasional buyers: .5(5-1) =.54 ≅.06. See also line B of Figure B.
3
89
503-019
A.
B.
C.
D.
E.
Customer Profitability and Lifetime Value
Occasional Buyers
Year One
Year Two
Year Three
Margin on each purchase
Survival rate
Cost of mailing catalogs
Total expected profit per customer
Cumulative profits per customer
(net of acquisition costs)
$16
100%
.5*12=$6
16-6=$10
$16
50%
.5*4=$2
.5*(16-2)=$7
$16
25%
$2
.25*(16-2)=$3.5
$(7.5)
$(0.5)
$3
From the analysis above, the direct retailer finds that frequent buyers become profitable by the end of year two. Occasional buyers become profitable by the end of year three.
Additional Comments:
a) In the above calculations, the discount rate has been ignored. The discount rate allows the seller to control for the fact that money received in the future is not equal to the value of money received today. Money today can be invested and interest on it can be earned. To adjust for discounting, total expected profits for each customer should be deflated for each a year by a factor of D = (1 + i)ª, where i is the interest rate, and a the number of years you wait to receive the money. The discount rate will be taken into account in the next part of the example. b) For simplicity, the analysis in line D assumes the retailer would cease sending catalogs to customers who drop out, in the same year they stop making purchases. In reality, a firm may send out the full number of planned catalogs prior to delisting a customer who does not make purchases in a given year. If this is the case, one should adjust line D and multiply the cost of mailing out catalogs by the survival rate of the preceding year (and not the current year). c)
In some cases, gross margins are not explicitly given and one has to establish them by incorporating all costs associated with a purchase (such as, delivery, payment to manufacturer, and processing fees). In other cases, the “net contribution” appearing on a company’s income statement may need to be adjusted for sales that eventually didn’t take place due to returns and cancellations (by multiplying contribution by (1-% sales returned)).
If net contribution includes costs of promotions and communications that are then separated out in the analysis, such as cost of mailing catalogs in the example above, those need to be excluded from the net contribution percentage.
Part III—Lifetime Value Analysis
Besides knowing how many years it will take to start making profits, it is obviously of interest for the catalog retailer to establish the total expected profit stream arising from each customer over the lifetime of his or her relationship with the company.
In the analysis below, we assume an interest rate of 10% for purposes of discounting future revenue streams. Expected profits are recognized at the end of each year.
4
90
Customer Profitability and Lifetime Value
Figure B
503-019
Lifetime Value Analysis for Frequent and Occasional Buyers
In line F, the cumulative profits per customer are determined in Year One by subtracting the acquisition costs
($17.5) from the net present value of profits. In subsequent years, the net present value of profits is added to the previous year’s cumulative profits per customer.
Frequent Buyers
A.
B.
C.
D.
E.
F.
Margin on each purchase
Survival rate
Cost of mailing catalogs
Total expected profit per customer Net present value of profits per customer (discount rate applied)
Cumulative profits per customer
(net of acquisition costs)
Occasional Buyers
A.
B.
C.
D.
E.
F.
Margin on each purchase
Survival rate
Cost of mailing catalogs
Total expected profit per customer Net present value of profits per customer (discount rate applied)
Cumulative profits per customer
(net of acquisition costs)
Year
One
Two
Three
Four
Five
Six
Seven
Eight
$10
100%
$6
$10
75%
$6
$10
56%
$6
$10
42%
$6
$10
32%
$6
$10
24%
$6
$10
18%
$6
$10
13%
$6
$14
$10.5
$7.9
$5.9
$4.4
$3.3
$2.5
$1.9
$12.7
$8.7
$5.9
$4
$2.8
$1.9
$1.3
$0.9
$(4.8)
$3.9
$9.8
$13.8
$16.6
$18.5
$19.8
$20.7
Year
One
Two
Three
Four
Five
$16
100%
$6
$16
50%
$2
$16
25%
$2
$16
12.5%
$2
$16
6.25%
$2
$10
$7
$3.5
$1.75
$.88
$9.1
$5.8
$2.6
$1.2
$0.5
$(8.4)
$(2.6)
$0
$1.2
$1.7
In computing lifetime value as depicted in Figure B, the retailer has assumed a relationship with the customer until the point where net present value of profits become less than $1 for the first time.4
This extends our previous break-even analysis across multiple years. As can be seen, a frequent buyer has an eight-year time span and generates a little over $20 of profits. An occasional buyer has an expected useful life of five years and generates about $2 in profits. The retailer may decide to send more (or fewer) catalogs or targeted promotions to these buyers to see whether they increase profitability or, alternatively, may decide they are not worthwhile pursuing and drop them once identified. To calculate aggregate profits, the retailer needs to know the size of each segment. This can be determined by the size of the pool from which customers will be acquired and the probability an acquired customer is a frequent/occasional buyer. Note that cumulative profitability of each customer differs from Figure A, given that discounting has been incorporated.
4 Clearly one could use other criteria, such as when the survival rate drops below some percentage. In some cases, the duration
of the relationship is known with greater accuracy in advance (such as reaching a certain age).
5
91
503-019
Customer Profitability and Lifetime Value
One can summarize the analysis presented in Figure B by formally writing an expression for customer lifetime value (CLV) as follows:
(M a − ca )r (a −1)
−
CLV =
(
1 + i )a a =1
N
1
∑
AC
where:
N = the number of years over which the relationship is calculated.
Ma = the margin the customer generates in year a. ca = the cost of marketing communications or promotions targeted to the customer in year a.
(a-1)
r = the retention rate (r
is the survival rate for year a).
i = the interest rate.
AC = the acquisition cost.
If Ma and ca are relatively fixed across periods, one can simplify the above expression by assuming an infinite economic life (i.e., letting N→∞), which leads to:
CLV2=
M −c
− AC
1− r + i
Using this approach, it is easy to establish that the CLV of frequent buyers is
20 − 6
− 17.5 = $22.5
1 − .75 + .1
Unless the retention rate is exceptionally high, this simplification produces results that are very close to the more precise formulation. The more years used in Figure B (or CLV1), the lower the discrepancy. Additional Comments:
a)
It is also possible to use a variation of CLV2 for occasional buyers. There is a need to adjust for year one because the cost of mailing catalogs is $6 instead of $2 for each subsequent year. This
16 − 2
⎛ 16 − 2 16 − 6 ⎞ yields −
− 17.5 − ⎜
⎟ = $ 2 .2
1 − .5 + .1
⎝ 1 + .1 1 + .1 ⎠
b)
If inflationary pressures are expected, one can adjust the expression for CLV1 by multiplying
(a-1)
the numerator by (1+u) , where u is the annual level of inflation. In CLV2, the denominator would become (1-r∗(1+u)+i).
c)
In some business contexts, the relevant time frame between purchases may be more than a single year (for example, with industrial clients upgrading computer hardware every three years). In these settings, calculating CLV may be easier when working with periods that recur every x number of years (x= time between purchases) than with annual values. The analysis would then require transforming the discount rate and inflation rate (that are typically given
6
92
Customer Profitability and Lifetime Value
503-019
in annual values), to span the duration of the period. Thus, in the expression for CLV1, a would be the number of periods and one would replace i by (1+i)x-1.
d)
In some business contexts Ma and/or ca, the per-period profits and costs of serving customers, may not be fixed over the entire lifetime of the customer. For example, as mechanical products age and require more maintenance, the servicing agent can expect to see increased profits from a customer. Annual average profits per customer may not highlight these changes. These make it more difficult to use the simplified expression CLV2. The same is true for the retention rate r, which may increase the more purchases an individual makes with the company. By contrast, competitor poaching and harsher economic conditions usually negatively impact the retention likelihood.
Strategic Implications of Customer Lifetime Value Analysis
The benefit from CLV calculations is two-fold: understanding the potential value of customers and prompting firms to learn more about the patterns of individuals or groups of customers. This information allows the firm to devise optimal strategies for each customer, eliminate wasteful costs, and create a long-term perspective of the potential relationship with customers. Firms can tailor strategies to deal with different customer segments that exhibit differences in buying characteristics at any given time, and they can also customize different strategies for the same customer depending on the stage of relationship between the customer and the firm.
Measuring CLV typically involves the use of historical sales data to allow detailed analysis of the profits customers are likely to produce across their “lifetime” (the time a company is likely to retain a customer). When a business is familiar with the historical buying patterns of its customers—and future customers are assumed to produce predictable levels of loyalty—then CLV can be used to estimate the likely profitability of those customers. As stated above, when a company has calculated
CLV, it is in a position to evaluate its customer base and devise marketing strategies for individuals or groups of customers. Common actions include:
a) “Firing” Customers
A company may find that only a fraction of its customers are profitable. Aside from attempting to make marginal customers become profitable, the company may be motivated to lose a percentage of the least-performing individuals. In some cases, this can be achieved by simply raising prices for the less profitable products. Best customers typically outspend others considerably, with a ratio of 15 to 1 in some industries. There is a competitive advantage for firms that can retain high value customers with minimal costs, and detailed information about customer profitability allows a firm to identify weaker customers. For purposes of firing customers, firms may also attempt to identify customers who create substantial costs, such as those making regular product returns or demanding increased levels of service.
b) Rewarding Customers
Likewise, a firm may choose to reward customers with discount vouchers or preferential services.
Identifying a firm’s best customers and investing in them can lead to higher loyalty (thus increasing their retention rate) and to improve sales.
7
93
503-019
Customer Profitability and Lifetime Value
c) Identifying Cross-Selling Opportunities
With detailed information about the interests and shopping patterns of individual customers, firms can identify opportunities to offer additional or related products (either separately or in the form of package deals). In some cases, firms may choose to proactively leverage collaborators whose products and competencies complement their own.
CLV and Relationship Marketing
In closing, it is worth emphasizing that the relationship between the customer and the firm typically evolves over time and is not static. To enhance the mutual value created by the relationship, it may be important for the firm to recognize the different phases of the customer life cycle. One subcategorization that can be useful for lifetime value analysis, suggested by Blattberg, Getz and Thomas in Customer Equity: building and managing relationships as valuable assets. (Boston, MA: Harvard
Business School Press, 2001), identifies five life cycle stages:
Stage 1: Prospects
Marketing effort is directed at attracting potential customers, who at this stage primarily develop expectations about the firm and its competencies. This stage is critical, therefore, for determining the long-term satisfaction and retention of a customer. If the firm creates expectations that are exceedingly high, customers may be easily acquired but will not be satisfied or retained, affecting both per-period revenues and survival rates.
Stage 2: First-Time Buyers
After making their first purchase, customers have some experience to assess the value created by the firm’s products and services for them—the risk of defection is typically high during this stage. For industries with short purchase cycles, satisfaction during this stage is critical for establishing repeat buyers. Stage 3: Early Repeat Buyers
Customers who make a repeat purchase are more likely to buy again than first-time buyers. As their confidence in the firm grows, their perception of value from the product increases and the rate of defection is reduced.
Stage 4: Core Customers
The rate of defection in this stage is typically low, and customers will not defect unless they encounter a high level of dissatisfaction.
Stage 5: Core Defectors
Many reasons, including the arrival of competitors with new products or enticing promotions and inconsistent delivery of service/quality by the firm, eventually cause some customers to switch suppliers. In some situations, external factors do not allow a firm to react to the loss of the customer,
8
94
Customer Profitability and Lifetime Value
503-019
while in other situations a firm can bring back a defector if the problem can be identified and rectified. Relationship marketing assumes the possibility of creating an ongoing relationship with a customer, and that customer satisfaction will be strong enough to create loyal or repeat buyers. In these situations, firms strive to find ways to increase customer satisfaction: provision of quality, additional services, and rewards for loyalty and repeat purchases. Satisfied customers tend to be less price sensitive and are likely to remain loyal for a longer period.
Evolving customer needs, competitors’ actions, opportunities to leverage collaborators, and changing economic and social conditions can all potentially affect the relationship between the firm and its customers. Customer lifetime value analysis, and the constant re-evaluation of the parameters determining it, presents the company with a framework to optimally respond to these changes.
References
Berger, P. and Nasr, N. “Customer Lifetime Value: Marketing Models and Applications.” Journal of
Interactive Marketing. Volume 12, No. 1 (1998).
Blattberg, R., Getz, G. and Thomas, J. Customer Equity: building and managing relationships as valuable assets. Boston, MA: Harvard Business School Press, 2001.
Blattberg, R. and Deighton, J. “Manage Marketing by the Customer Equity Test.” Harvard Business
Review (July-August, 1996).
Dwyer, R. “Customer Lifetime Valuation to Support Marketing Decision Making.” Journal of Direct
Marketing. Volume 11, No. 4 (1997). th Kotler, P. Marketing Management: analysis, planning, implementation, and control. 9 Edition. Upper
Saddle River, NJ: Prentice Hall, 1997. th Kotler, P. and Armstrong, G. Principles of Marketing. 9 Edition. Upper Saddle River, NJ: Prentice Hall,
2001.
Rust, R., Zeithaml, V., and Lemon, K. Driving customer equity: how customer lifetime value is reshaping corporate strategy. New York: Free Press, 2000.
Schnaars, S. Marketing Strategy: A Customer-Driven Approach. New York: Free Press, 1991.
9
95
96
9-500-024
REV: JANUARY 15, 2002
JOHN DEIGHTON
The Brita Products Company
In 1987 when Charlie Couric saw his first Brita pitcher he thought, “A homemade alternative to bottled water!” Here was a product that, with the right marketing support, could be very successful.
Couric, a marketing executive with the Clorox Company charged with finding new business ideas, had been browsing health food stores in California when he came across the quirky home water pitcher-and-filter system made by a small German company, Brita GmbH. He proposed that Clorox acquire the right to market Brita in the USA, and in 1988 they did so. Couric reflected:
In the early years we had to beg the corporation to invest. Some of my colleagues viewed the pitcher as another waffle iron - used once and then tossed into the basement. We saw it differently. We looked at the repeat purchasing of filters, and to us the strategy was obvious.
This was a race to put a pitcher on every kitchen countertop, at a loss if necessary.
Clorox supported Couric’s deficit-spending proposal, and a decade later Brita had grown to become one of Clorox’s biggest brands. It had rewarded Couric’s faith, spearheading the growth of a home water filtration industry in the United States. More than 17 million Brita pitchers had been sold, and each pitcher sale started a stream of filter sales. The Brita brand was generating close to
$200 million revenues a year.
Now, in 1999, Couric was keeping an eye on a different water purification product launched by a small competitor, PUR. It was a filter that screwed onto kitchen faucets.1 Clorox had its own version of the faucet-mounted filter ready for launch, and again a debate had developed over whether to deficit-spend. Some counseled that the faucet-mount had the power to disrupt the pitcher product, and Brita had no choice but to pour money into another race to build another installed base, this time in faucet-mounted products. Others argued that faucet-mounts served a different niche of the water purification market from pitchers, and the two could live side-by side. A third group argued that
Brita should do nothing to foster faucet-mounts. Its priority, they argued, was to invest to defend its installed base of pitchers and the associated filter revenue stream. Money spent on promoting a faucet-mount would only erode the pitcher base and interrupt its stream of filter revenues. They pointed out that PUR was a small, loss-making firm, too weak to succeed at creating a new category, particularly when the early adopters of home water filtration all had pitchers in their homes.
1 Faucet-mounted filters were not themselves a new product—Teledyne had sold one since the 1960s without much success.
However, the introduction of solid carbon block technology in 1995 improved performance and increased consumer interest.
________________________________________________________________________________________________________________
Professor John Deighton prepared this case. HBS cases are developed solely as the basis for class discussion. Cases are not intended to serve as endorsements, sources of primary data, or illustrations of effective or ineffective management.
Copyright © 1999 President and Fellows of Harvard College. To order copies or request permission to reproduce materials, call 1-800-545-7685, write Harvard Business School Publishing, Boston, MA 02163, or go to http://www.hbsp.harvard.edu. No part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the permission of Harvard Business School.
97
500-024
The Brita Products Company
The Clorox Company
The Clorox Company was a major manufacturer and marketer of laundry additives, household cleaners, charcoal, auto care products, cat litter and home water purifiers. In January 1999, Clorox bought First Brands, a $1.2 billion manufacturer of plastic wraps and bags, auto care products, cat litter and home fireplace products. Revenues of the two companies combined would have been $3.9 billion in 1998. Some of the well-known U.S. consumer brands that would come under common ownership following the merger were:
Clorox
First Brands
Armor All car care products
STP automotive products
Fresh Step cat litter
Scoop Away cat litter
S.O.S. steel wool pads
Ever Clean cat litter
Hidden Valley salad dressing
Johnny Cat cat litter
Kingsford charcoal
StarterLogg fire starters
Clorox laundry bleach
HearthLogg fire logs
Soft Scrub cleaners
Glad plastic wraps and bags
Brita water filtration systems
Formula 409 spray cleaner
Tilex cleaners
Pine-Sol cleaner
Liquid Plumr
Under Chairman and CEO G. Craig Sullivan, Clorox followed a strategy of building dominant brands, pursuing international expansion and acquiring promising businesses. Some 85% of Clorox brands were first or second in their categories. Sales beyond the United States would reach 20% of revenues after the First Brands merger, up from 18% for Clorox alone.
The Brita Products Company
Brita GmbH, a family-owned corporation headquartered in Tanusstein, Germany, made a variety of industrial and consumer water filtration products. Before Couric called, it had struggled without much success to sell its home water filtration system in the United States, most recently through a
Canadian agent. After vigorous negotiation, in September 1988 it agreed to let Clorox form a subsidiary, Brita USA, to be the sole U.S. distributor of Brita products. Clorox would buy filters from
Brita GmbH and design and make its own pitchers. Couric became President and General Manager of Brita USA.
For four years, Brita USA struggled. The costs of building distribution, designing products, and promoting the concept dwarfed the small base of sales. Couric persevered, however, because early surveys of users suggested to him that a Brita customer would have a remarkable lifetime value.
Each pitcher sale would start a flow of filter sales. Over 80% of pitcher buyers were still using the product a year later and many were telling friends to try it and were giving it as gifts. In the 1990s, the product took hold like crabgrass. By 1999, an estimated 13% to 15% of the 103 million households in the United States were using a Brita pitcher. Brita had created a home water purification industry worth $350 million at retail, and held a 70% revenue share.
2
98
The Brita Products Company
500-024
The Product
The Brita pour-through filtration system comprised a two-compartment pitcher and a replaceable filter (Exhibit 1). Tap water, poured into the upper compartment, flowed under gravity through the filter into the lower compartment, filling it in about five minutes.
The filter had two elements. Activated carbon reduced chlorine, sediment and odors, and an ionexchange resin removed any heavy metals such as lead, copper, mercury and cadmium, as well as temporary water hardness (calcium and magnesium). The benefits were threefold. Filtered water tasted better, it did not deposit scale when boiled, and, to the extent that it might have contained harmful heavy metals, they were extracted. The filter did not screen out microorganisms such as cryptosporidium and giardia, two sources of gastro-intestinal illness that were potentially fatal to people with compromised immune systems.
The pitcher system was sold with a single filter in place. Filters required replacement every two months or after filtering 40 gallons of water. Brita supplied calendar stickers to help users track when a filter needed replacement. Filters were sold in packs of one, three and five.
Consumer Attitudes and Behavior
Over the decade of the 1990s, the safety of tap water became a topic of growing concern to U.S. households. In one well-publicized case, two wells supplying the drinking water for the Boston suburb of East Woburn, MA were found to be contaminated with industrial solvents, coincident with a number of cases of leukemia. The incident was the subject of a book and a 1998 motion picture, ‘A
Civil Action.’ In that year the U.S. Environmental Protection Agency declared that about 10% of the sediment under U.S. surface waters is "sufficiently contaminated" with toxic pollutants to pose a health threat to humans and wildlife. Later that year, Congress began requiring municipal water authorities to say when contaminant levels exceeded federal regulations. In Milwaukee in 1993,
403,000 people were made sick and 111 died when the parasite cryptosporidium entered the municipal water supply. By the end of the decade, a poll by USA TODAY, CNN and Gallup found
2
that 47% of respondents preferred not to drink water straight from the tap.
Sales of bottled water from U.S. supermarkets and home delivery services grew rapidly during the decade. By 1997, bottled water made up 8% of all the liquids that people paid to drink and was the industry’s fastest growing category (Exhibit 2). In that year, 45% of households bought still (noncarbonated) water in supermarkets and 27% bought carbonated water. The average price paid for still water was about a dollar per 128-ounce container, and carbonated brands averaged about three dollars (Exhibit 3).
3
A 1998 survey found that two-thirds of Americans claimed to be familiar with the expert recommendation to drink eight, eight-ounce servings of water a day, yet only one in five drank that quantity of water and 44 percent drank three or fewer water servings daily.
2 Eisler, Peter, Barbara Hansen and Aaron Davis. “Lax oversight raises tap water risks.” USA TODAY, October 21, 1998, p.
15A.
3 Yankelovich Partners conducted a survey of 3,003 Americans for the Nutrition Information Center at The New York Hospital
- Cornell Medical Center and the International Bottled Water Association. The study is described at http://www.bottledwaterweb.com/news/news3.html. 3
99
500-024
The Brita Products Company
A 1999 survey4 found that 72% of all respondents, and 89% of young adults, voiced some concern about the quality of their household’s water supply. A majority of households used either bottled water or some water purification system to limit their exposure to public water supplies. The number taking no precautions declined from 47% in 1995 to 35% in 1999 (Exhibit 4).
Market Performance
Brita used the term ‘systems’ to refer to pitchers and faucet-mounted units, and ‘filters’ for the replacement filters. System sales were sluggish for the first four years after launch, but filter sales grew more rapidly (Exhibit 5). In the early years Couric compared performance in the United States to the first years of the product’s life under other Brita distributors in Canada and the United
Kingdom, and the similarity of the profiles gave him, as early as 1991, the confidence to persevere:
We saw that trial in the United States in the early 1990s was running between the Canadian and the U.K. levels. Close to 25% of buyers told us that they had given a Brita pitcher as a gift.
Another reassuring sign was that surveys were finding that more than 80% of those who had tried the pitcher were still using it a year later. The same surveys reported that they were buying 2 or 3 filters a year. Each year we tried to relate filter sales to past pitcher sales. We found that when we estimated our installed base at 80% of those who had bought a system in the previous five years, and assumed that the installed base bought 2.5 filters a year, the resulting forecasts of filter sales each year were close to reality.
Management tracked system market share in units and filter share in dollars (Exhibit 6). Brita’s share of combined system and filter market revenues had been steady in the range of 65% to 75% from 1995 to 1999. System unit shares were far more volatile. In July 1998, for example, system sales doubled over the previous year and Brita’s share increased ten points in response to a so-called
‘bogo’ (buy one, get one free) promotion on pitchers, intended to pre-empt a PUR competitive launch.
Filter sales had been less responsive to sales promotion activity by manufacturers.
Brita bought pitchers from contract manufacturers at a cost per unit of $7.80. Filters were purchased from a manufacturing plant owned jointly by Clorox and Brita GmbH for $2.05 per unit, inclusive of a 3% royalty to Brita GmbH. (Profits from this plant were not material.) At the prices
Brita charged retailers in 1999, pitchers earned a contribution to fixed costs of 48.6% of Brita’s net revenue, and filters earned 50.0%. After advertising and trade promotions, Brita USA earned a net return on sales of 24%, the highest of all Clorox business units. Although advertising spending worked to the benefit of both pitchers and filters, trade promotions were used mainly to secure trade support for pitchers. Exhibit 7 summarizes the income statement of the brand in 1998.
Distribution
At inception, Brita’s main retail outlet had been a health foods chain. Its competitors were in stores that stocked housewares, like Sears and Walmart. Couric believed, however, that the product would flourish in Clorox’s traditional base, grocery and drug outlets, and drove distribution in that direction. 4 “1999 National Consumer Water Quality Report,” Lisle, IL: Water Quality Association, 1999.
4
100
The Brita Products Company
500-024
Brita distribution
1992
1998
Department stores
27%
13%
Mass merchandisers
31%
34%
Grocery stores
11%
14%
Club stores
31%
21%
Drug stores
-
12%
This pattern of channel evolution, in which high margin retailers like department stores pioneered a new category, only to lose share to lower margin retailers, was known among marketers generally as a “class to mass” strategy. Couric explained his strategy:
Our version of “class to mass” had three elements. We wanted to be established in class, first in mass, and alone in grocery. So we created a line of upscale pitchers called Ultra for department stores, appropriate for their 35% mark-up structure. We sold the standard pitcher, inherited from Germany but manufactured locally, in mass merchants like Target and Walmart and in drug and grocery stores. They marked it up 25%. We designed a bonus pack system and a 5-pack of filters to appeal to club retailers when they became important in the early
1990s.
Keeping these various classes of trade happy is an enormous challenge. We are continuously building the system and seeing cracks appear. Eventually perhaps we’ll be driven out of ‘class,’ but we aim to make it last as long as possible. We need the breadth of distribution and variety of products to support our $30 million advertising budget, and to provide a channel for introducing and establishing future new products.
One way we attempt to keep the peace among our classes of trade is by insisting that no retailer advertise a Brita line at below the price that we set. We call it a MAP
(minimum advertised price) policy. We reimburse retailers for featuring our products in their display advertising, but if they feature us at below the MAP, then we won’t pay. The only exception we make is on the standard pitcher. We let them deal as much as they like on that item.
Positioning and Advertising
Brita’s advertising in the United States emphasized a taste benefit. Couric explained:
Initially people had no clue about the concept of a pitcher product. I remember on the cab ride back to the airport after our first trade show in Chicago I explained the concept in lengthy detail to the cab driver and when I was finished he said, ‘You screw it onto the faucet, right?” I realized then that we had to tell people how the product worked, so we split the advertising message 50/50 between how it works and how it tastes. Today, now that our product is more well known, we are able to be more focused about the taste benefit.
We decided on taste for three reasons. First, research showed that when we talk taste, we get a health benefits halo. When we talk health, we don’t communicate the idea of taste.
Second, we noted that the whole bottled water industry had been built without reference to health. Third, I wanted to be at the top of the mountain. I didn't want competitors overtaking us. If we focussed on lead removal, say at 93%, someone else could claim to take out 95%.
5
101
500-024
The Brita Products Company
With taste, we could say it first, say it loudest, and we could own the benefit. By now, with
$100 million of cumulative advertising on the taste claim behind us, we are impossible to dislodge. When we started eleven years ago, the water filtration category had low credibility. It was being investigated all the time for improper or false claims. We didn’t want to get into a claims war. Our advertising needed to be pure and simple. We showed mountain streams, waterfalls and the outdoors. We promised clear, crisp, refreshing water, which is what we delivered.
Today we own the waterfall imagery.
Competition
Brita’s success attracted competitors in droves. Among the brand names that entered the market were Culligan, Electrolux, Sunbeam, Kenwood, Corning, Melitta, PUR, Rubbermaid, Teledyne, Omni and Mr. Coffee. None had succeeded in challenging Brita’s leadership, which remained in the range of 65% to 75% across the decade. In 1999, the only competitor with double digit market share was
PUR, the brand of a small, publicly held U.S. corporation, Recovery Engineering. This company made, in addition to water filters, a line of portable drinking water systems for outdoor enthusiasts and desalinators for marine and military use.
At the January 1998 International Housewares Show in Chicago, a dozen manufacturers unveiled new water filtration products or extensions to existing lines. The products appeared to be designed to attack niches not currently served by Brita. "Number 1 is always going to be under attack," Brian
Barton, brand manager at Rubbermaid, told the press covering the trade show. "When you have 80%
5
of the market share there's only one way to go and that's south."
Several products took health and safety positions in reaction to Brita's taste appeal. Number 2 brand, PUR, announced that it would spend $40 million in advertising and promotion to support its line of faucet-mount and pitcher filters. The spend included a $15 million outlay for PUR Plus, a new pitcher system touted as the most technologically-advanced to date. The PUR filter would remove contaminants such as cryptosporidium and giardia. PUR representatives described a promotional program that would begin with a six month infomercial running on national cable television, to be followed by a schedule of 30-second spots on cable and spot network TV that would point out the differences between PUR and current pitchers. Sunbeam, known for blenders and toasters, announced the launch of Fresh Source, a product that removed microbiological cysts as well as chlorine and lead. Sunbeam would back the product with an estimated $10 million of advertising.
Number 3 manufacturer Teledyne, unveiled a faucet-mount product at the show.
At the trade show, Brian Sullivan, president and CEO of Recovery Engineering, Minneapolis was quoted as saying, "If you ask consumers if they want more contaminants taken out or less, they'll say more. People will pay more for a higher-performing product." PUR’s pitcher was considerably more expensive than Brita’s standard product. A Brita representative responded, "The way we see the market, this business is geared more towards taste. Consumers are interested in taste. Bacteria is way
6
down the list."
In a $30 million ad campaign unveiled in the month of the Chicago show, Brita did not mention harmful impurities. A TV spot dubbed "There Was a Time" features shots of rushing streams set
5 Mehegan, Sean. “Sunbeam, recovery loading up $$ to take on Brita in water filtration.” Brandweek, Vol. 39, Iss. 3, January 19,
1998 p. 12.
6 Op cit., p. 12.
6
102
The Brita Products Company
500-024
against a backdrop of mountains and a dark, brooding sky. "There was a time when it was perfect," the voiceover says. "You can have this taste . . . again."
Rubbermaid had launched a low priced product in 1997, which used a technology similar to
Brita’s while attacking it on price/performance. Rubbermaid claimed that its filter could cleanse 800
8-oz. glasses of water versus 560 for Brita at the same price. In 1998 the company repackaged its pitcher product and announced a portable 16-oz. bottle with a carbon filter built into the cap.
Rubbermaid had not advertised its pitcher in 1997, and sales had been disappointing. At the trade show they pledged to step up promotion under a new team led by Cathryn Rings, the former head of
Procter & Gamble's Max Factor cosmetics business. Rings announced at the trade show, "We're going to try some classic P&G marketing.”
The Faucet Mounted Filter Entry
Prior to 1995, Brita executives expressed little interest in faucet mounted water filtration systems.
Of the 50 countries in which Brita GmbH did business, only Japan had a significant faucet business, and that was attributed to space constraints in Japanese kitchens. In 1994 and 1995, however, they saw a faucet segment forming. Recovery Engineering Inc. launched a faucet mounted product under the PUR brand name with some success. In 1995, Brita hired an outside design company to design a faucet mount.
Functionally, water from a pitcher was different from water filtered through a faucet-mounted filter. In favor of pitchers, they were usually stored in refrigerators, so pitcher water was cold while faucet water was not. In those parts of the country where tap water was ‘hard’ and left scale deposits and scum when boiled, only pitcher filters but not faucet filters would eliminate hardness. In favor of faucet mounts, the water passed through at higher pressure than through pitcher filters, so finer filters could be used that could screen for microorganisms and offer protection against cryptosporidium and giardia. Also, pitcher-filtered water tasted crisper, with lower pH. Finally, faucet-filtered water cost significantly less per glass because the filter lasted longer. Where pitcher water cost 15 to 20 cents per gallon, faucet-filtered water cost perhaps half.
Were these differences significant to consumers? Did Brita stand for good tasting water, or how you get it? The Brita team debated whether the faucet mount would be perceived as another way to deliver Brita water. Or would the consumer decide that they were buying something quite different, perhaps even so different that some might consider it a good idea to own both a pitcher and a faucet mount? As Brita’s filtration technology played no part in the faucet mount design, Clorox was not obliged to use the Brita name on this product. If it did so, however, it was required to pay Brita GmbH a royalty that, under the 1988 agreement, would be between 3% and 4% of sales depending on the magnitude of sales. It would also be bound by the non-compete clause of the comprehensive agreement that limited sales of products with the Brita name to North America. Conversations with the retail trade, however, revealed a distinct preference for carrying the faucet mount under a wellknown name like Brita.
The direct cost of the faucet mount system was estimated to be $15.00 and the direct cost of a replacement faucet filter would be $3.00. Pitcher filters could not be used in faucet mounts.
Would a faucet mount cannibalize pitcher and pitcher filter sales? Perhaps, some speculated, the pitcher was a starter product, and customers who had learned to go back to drinking tap water would graduate up to the more convenient and sophisticated faucet unit. To explore these and other
7
103
500-024
The Brita Products Company
7
questions, Clorox commissioned a simulated test market from ACNielsen Vantis, a division of the
ACNielsen BASES group.
Market Simulation Study for the Faucet Filter
8
In the ACNielsen Vantis study, 567 respondents, characterized as water-involved and drawn from eight markets across the United States, were intercepted in shopping malls, brought to rooms containing simulated retail store shelves, and asked to choose from a display of water filtration systems. Respondents were assigned to one of three rooms. Each room displayed ten products currently available in the market. In addition, two displayed a prototype of the Brita Faucet Filter System, one priced at $34.99 and the other at $39.99. The third had no Brita Faucet Filter System on display, to serve as the control cell of the experiment. In the first two rooms, subjects saw print advertising for
Brita and PUR faucet mounted filters.
Consumers were first asked to rate how likely it was that they would buy any of the displayed items in the next two months. They were then asked to identify their first, second and third choice of item. Finally, they were asked a series of questions about the test product, the Brita Faucet Filter
System. Vantis offered the Brita team the following conclusions from the study:
•
The Faucet Filter increased the likelihood of buying a product from the Brita line.
•
However it did not increase interest in the filtration category as a whole, so that the combined pitcher and faucet-mounted market was not expected to expand.
•
Though higher priced, the Brita Faucet Filter generated similar levels of purchase intention to the Brita Spacesaver Pitcher.
•
About half the Brita pitcher owners who bought the Faucet Filter system would continue to use the pitcher in conjunction with the faucet product.
•
Both Brita and PUR’s faucet filters were perceived to be superior to the Brita pitcher in removing contaminants, and in convenience. However, only Brita’s faucet filter was perceived to improve the water’s taste.
•
Unit sales and perceptions of value for the faucet mount were strong at both the $39.99 and
$34.99 prices, and sales would not be significantly impacted if the PUR price was dropped by
$5.00.
7 Simulated test market studies have a long history, dating back to the 1970s when Alvin J. Silk and Glen Urban began research
at MIT’s Sloan School to seek ways to forecast demand for new products without incurring the costs and public exposure of full-scale test markets. Today simulated test markets are regularly relied on to forecast in-market performance without the need to build production capacity, expose marketing plans to competitive scrutiny, and wait six or twelve months to read results. Their evolution is described in Kevin J. Clancy, Robert S. Shulman and Marianne Wolf, Simulated Test Marketing:
Technology for Launching Successful New Products, New York, NY: Lexington Books, 1994. ACNielsen Vantis serves services and durable goods industries with a brand of simulated test market methodology, BASES, that derives from work that began in the
Pillsbury Company in the 1960s, found a temporary home in Booz-Allen & Hamilton (the name is an acronym for Booz-Allen
Sales Estimation System), was spun off in a leveraged buyout in 1977, and is now a division of the A. C. Nielsen Company.
8 A water-involved respondent was one who either owned a filtration device or bought bottled water, and described him- or
herself as not satisfied with the quality of their water.
8
104
The Brita Products Company
500-024
The study projected that unit sales of the faucet mount in its first year would lie between 350,000 and 1,395,000 units. Half of this volume would come from consumers who would otherwise have bought a Brita pitcher. Whether sales would be at the high or low end of the range depended on how aggressive was Brita’s marketing investment, and how competitors responded. Ten scenarios were generated by combining the following factors and levels:
Low
Consumer advertising
Consumer promotion
Feature price reductions
Other trade spending (displays, racks, etc.)
$5.4 million
$2.0 million
$1.8 million
$3.2 million
High
$11.1 million
$3.0 million
$2.3 million
$6.1 million
Very High
$15 million
$4 million
$2.8 million
$9 million
The ten scenarios gave rise to the following ten sales forecasts:
Scenario:
1
2
3
4
5
6
7
8
9
10
Minimum
Advertised Price
($5.00 off List)
$34.99
$34.99
$29.99
$34.99
$29.99
$29.99
$34.99
$34.99
$29.99
$29.99
Consumer
Promotion and
Trade Spending
Consumer
Advertising
Competitive
Pricing
Low
Low
Low
High
High
High
High
Very High
Very High
Very High
Low
Low
Low
High
High
High
Very High
Very High
High
Very High
Low
Current
Low
Low
Low
Current
Current
Current
Current
Low
First Year Unit
Sales Forecast
340,000
350,000
395,000
970,000
1,125,000
1,160,000
1,205,000
1,245,000
1,350,000
1,395,000
Each scenario resulted in a sales and income forecast. For example, scenario 2 led to the following forecast: Total households
Product awareness resulting from $5.4 million advertising
Distribution (% of market reached)
List price (30% of sales assumed to occur at list)
Feature price (70% of sales assumed to occur at the feature price)
Trade promotion
Consumer promotion
Competitive pricing
Total sales (units)
75.86 million
13%
72%
$39.99
$34.99
$3.2 million (low)
$2.0 million (low) current levels
350,000
Brita USA had not asked Vantis for a forecast of sales of replacement filters. The proposed product had an LED filter replacement indicator, which would likely increase compliance with filter replacement recommendations. Each filter would treat 100 gallons of water, about four months’ output from a typical kitchen faucet, before the indicator would signal that it was due for replacement. 9
105
500-024
The Brita Products Company
Couric’s Decision
Couric prepared to call his marketing team together to hear their views on how to take the Brita brand forward. He anticipated that he would hear three points of view: keep the focus on building the installed base of pitchers, shift the budget to encourage the installed base to buy more filters, or put the weight of resources behind building a whole new installed base in faucet-mounts.
He saw many demands on the Brita marketing budget besides the faucet-mount product launch.
Household pitcher penetration was slowing, and yet six out of seven households did not have one.
Could there be segments who had not responded to the broad appeals of the first decade, but who might well respond to more targeted communication efforts—specific appeals to singles and to parents of young children, for example? Perhaps investing in direct mail or other highly targeted marketing tools could cultivate demand in these niches. Then there was the filter opportunity. Brita had never invested in the direct cultivation of filter demand, beyond in-store promotion.
On his desk in the corner office were Recovery Engineering’s published financial results for the quarter ending January 3, 1999. Its quarterly sales were up $1 million on the previous quarter to
$19.5 million, but its net loss had more than doubled to $7.2 million. Its stock was trading at $10, down from $35 in mid 1998. Recovery Engineering had raised capital in an Initial Public Offering in
1997 on the claim that it had a technological edge over Brita. To be sure, PUR had been first to market with a number of new features: the first cryptosporum filter for pitchers, the first mechanical device to indicate when to replace a filter, and the first widely distributed faucet filter. But as Couric weighed how much urgency to put behind the faucet-mount launch, he found comfort in PUR’s flow of red ink.
A Fax and a Phone Call
Couric’s fax began transmitting. Over the line came a report from the Clorox field sales director, with a sketch of a display that had been seen that morning in the Schaumburg, IL branch of Target
Stores, in the store’s large water filtration section. A sign over the display had read:
Which water filtration product is right for me?
How do different water filters give me great tasting water and protect my family?
Choose your level of protection:
•
••
•••
••••
Lead, chlorine.
Lead, chlorine, cryptosporidium, guardia.
Lead, chlorine, cryptosporidium, guardia, Lindane (a pesticide), Atrazine (a herbicide), asbestos.
Lead, chlorine, cryptosporidium, guardia, Lindane (a pesticide), Atrazine (a herbicide), asbestos, benzene, TTHMC.
All product claims are NSF® certified to national public health standards.
Beneath the sign he had seen five PUR systems and four Brita systems mounted on identical backing cards labeled with one, two, three or four bullets. No PUR system had fewer than two bullets and the PUR Ultimate Faucet Mount had the maximum, four. Not a single Brita system had more than one bullet.
Simultaneously, Couric’s phone rang. His investment bankers were advising that Procter and
Gamble, the world’s largest consumer products company and Clorox’s most respected competitor, was about to close a deal at $35 per share for control of PUR.
10
106
The Brita Products Company
Exhibit 1
500-024
1993 Advertising
11
107
108
Exhibit 1
(continued)
Brita offers a money-back guarantee (details in box). Brita is available at many fine retailers. For the one nearest you, call 1-800-44-BRITA, or visit our website at www.brita.com. Substances removed may not be in all users’ water. ©1997 The Brita Products Company.
The ingenious filter inside the Brita® Water Filtration Pitcher is one of a kind.
It virtually eliminates lead and chlorine. Dramatically reduces copper, sediment and water hardness. Best of all, it turns tap water into clear, fresh, wonderful water. There’s nothing like the great taste of Brita water. Enjoy.
500-024
-12-
The Brita Products Company
Exhibit 2
500-024
Growth of Segments of the U.S. Beverage Market
Cumulative Growth (Indexed on 1988)
Bottled Water
160%
Fruit Juice
130%
Beer
Soft Drinks
Total Market
Tea
100%
Milk
Coffee
Wine
Spirits
70%
1988
1990
1992
1994
1996
Bottled
Water
Fruit
Juice
Soft
Drinks
Beer
Wine
Per Capita
Consumption
(gallons, 1997)
12.7
15.0
54.6
29.1
1.9
1.2
Segment Size
(Billions of gallons, 1997)
3.4
4.1
14.7
7.9
0.5
0.3
Spirits
Coffee
Tea
Milk
22.6
7.6
20.0
6.1
2.1
5.4
Source: 1999 Beverage Marketing Directory, Mingo Junction, Ohio: Beverage Marketing Corp., 1999.
Total Per Capita Consumption = 144.7 gallons in 1997.
Total Beverage Consumption = 44,500,000,000 gallons in 1997.
13
109
500-024
The Brita Products Company
Exhibit 3
Major Brands in the Bottled Water Category (supermarkets only)
Percent of
Households
Buying
Price per 128 oz. unit
Still Water Brands
Dannon
Arrowhead
Poland Spring
Sparkletts
Chrystal Geyser
Evian
Hinckley & Schmitt
Private label
44.77%
6.27%
4.57%
4.47%
4.05%
3.85%
2.61%
2.38%
17.73%
$1.03
$2.09
$0.95
$1.39
$0.86
$2.13
$5.49
$1.36
$0.68
Carbonated Water Brands
Canada Dry
Schweppes
Vintage
Clearly Canadian
Perrier
Private Label
26.83%
5.69%
4.66%
2.65%
1.75%
1.31%
12.52%
$3.70
$4.97
$5.85
$2.44
$11.66
$9.47
$2.78
Source:
Information Resources Inc. “Marketing Fact Book” January-December 1997,
http://fic.wharton.upenn.edu/iri/factbook .
14
110
The Brita Products Company
Exhibit 4
500-024
Water Quality: Consumer Attitudes and Behavior
From a survey of 1,007 adults conducted between January 14 and 17, 1999. Sample is projectable to all U.S. adults over age 18.
Expressed concerns about household water quality (% of respondents):
Expressed any concern
Health contaminants (net)
Bacteria
Aesthetics (net)
Smell and/or taste
Appearance
Hardness
Sediments
1995
1997
1999
75
54
75
50
61
45
33
38
43
60
42
26
39
40
72
48
9
56
40
26
36
38
Expressed concerns about household water quality by age in 1999 (% of respondents):
Expressed any concerns
Health contaminants
18-24
25-34
35-44
45-54
55-64
89
67
79
56
79
58
70
45
59
36
65+
56
25
Use of water treatment device (% of respondents):
No device used
Bottled water
System
Table-top pitcher
System on faucet
Whole house system
Softener
1995
1997
1999
47
36
27
5
9
10
10
41
37
32
12
11
7
9
35
38
38
16
11
8
9
Use of water treatment device by region of country in 1999 (% of respondents):
North East
Bottled water
Table-top pitcher
System on faucet
Whole house system
Softener
Source:
33
24
9
10
5
North Central
32
12
10
8
21
South
West
42
17
10
7
4
43
12
14
10
6
Water Quality Association: 1999 National Consumer Water Quality Report.
15
111
500-024
The Brita Products Company
Exhibit 5
Brita Unit Sales, 1989 to 1998
Brita Unit Sales (’00)
Systems
Filters
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
171
402
194
581
202
876
302
1292
546
2,205
1,056
4,458
2,030
8,164
3,363
15,246
4,565
23,293
5,266
27,413
Sales of Brita Pitcher Systems and Filters
30000000
25000000
Unit Sales
20000000
Systems
Filters
15000000
10000000
5000000
0
1988
Source:
1989
1990
1991
1992
1993
1994
Company records (approximate and unaudited).
16
112
1995
1996
1997
1998
1999
The Brita Products Company
Exhibit 6
500-024
Retail Market Shares (United States, all retail outlets)
Systems
1992
Pitchers (thousands of units)
Brita
PUR
Rubbermaid
All others
Faucet Mounts (thousands of units)
PUR
Teledyne
375
82%
18%
1,186
23%
1993
640
82%
18%
782
23%
1994
1995
1996
1997
1998
1,405
75%
25%
602
30%
2,636
77%
23%
659
9%
43%
4,381
77%
23%
898
30%
43%
5,689
80%
4%
7%
11%
1,249
67%
27%
6,307
83%
8%
4%
5%
1,291
74%
23%
Filters
Filter sales ($millions, retail)
Brita
Teledyne
PUR
Omni
Sears
Pollonex
Source:
1992
1993
1994
1995
1996
1997
1998
$20.5
32%
25%
0%
12%
7%
7%
$26.5
43%
20%
0%
13%
6%
4%
$38.7
59%
15%
0%
8%
3%
2%
$63.3
65%
10%
1%
8%
2%
1%
$82.3
75%
9%
2%
5%
2%
1%
$116.3
75%
7%
8%
3%
1%
0%
$154.7
75%
4%
17%
2%
1%
1%
Company records, assembled from data from Industrial Market Research Inc. and Information Resources, Inc.
Note: Retail audit estimates may not agree precisely with Brita’s sales records because they are extrapolated from a sampling of retail stores.
17
113
500-024
Exhibit 7
The Brita Products Company
Revenue and Net Income of Brita Systems and Filters, 1998
Total (‘000)
Per Unit
Brita Pitcher Systems
Unit sales
Revenues
Cost of goods sold
Gross margin
Consumer promotion
Feature price reductions
Other trade spending
5,266
$79,800
41,100
38,700
4,000
5,000
7,000
$15.16
7.80
7.36
Brita Filters
Unit sales
Revenues
Cost of goods sold
Gross margin
Consumer promotion
Feature price reductions
Other trade spending
27,413
$112,400
56,200
56,200
1,000
1,000
1,000
$4.10
2.05
2.05
Combined Brita Systems and Filters
Combined revenues
Combined gross margins
Advertising
Combined consumer and trade promotions
Net income before G & A
$192,000
94,900
30,000
19,000
45,900
Source: From company records, modified to preserve confidentiality of margin information but without altering the relative magnitudes of margins.
18
114
UVA-M-0451
Rev. Apr. 12, 2013
THE INDEPENDENT ADVISER FOR VANGUARD INVESTORS
Although Dan Silver’s MBA education was backed by considerable experience in editing and publishing, he felt challenged by some of the details of his new job as director of marketing and operations for the Independent Adviser for Vanguard Investors (TIAVI), a financial newsletter. In particular, many of the complexities surrounding the renting of mailing lists and the direct-mail solicitation of additional customers were new to him. Currently there was growing pressure within his company to cut back on prospecting because several recent efforts had lost money. It was clearly his responsibility to respond to the pressure.
Newsletter Background
Dan Wiener, a former U.S. News & World Report associate editor, started the Vanguard
Adviser in 1991. The newsletter’s purpose was to give advice to investors in the Vanguard Group family of mutual funds and to make money doing so. In 1993, the Vanguard Group was the second-largest group of mutual funds, managing over $100 billion of assets in its more than 70 funds. Approximately 2.5 million individual investors had money invested in one or more
Vanguard funds. For many of these investors, the advice of a Wall Street outsider like Wiener was seen as particularly valuable. Not only was Wiener not associated with Wall Street, he had no ties to the funds about which he gave advice. Because he had nothing to sell, Wiener’s independent advice was seen by subscribers as more valid and trustworthy. In addition, mutualfund companies such as the Vanguard Group were prohibited by law from giving advice on their own funds.
In early 1992, Wiener and his company, the Fund Family Shareholder Association, had brought in two partners from the publishing business to manage the day-to-day operations of the newsletter. Wiener, of course, remained the president. Later that summer, the Fund Family
Shareholder Association hired Silver as director of Marketing and Operations. In May 1993, the newsletter made Marketing Services International, Inc., (MSI) the manager of its member and inquiries file.
This case was prepared by Professor Phillip E. Pfeifer. It was written as a basis for class discussion rather than to illustrate effective or ineffective handling of an administrative situation. Some elements in the case are disguised.
Copyright ¤ 1994 by the University of Virginia Darden School Foundation, Charlottesville, VA. All rights reserved.
To order copies, send an e-mail to sales@dardenbusinesspublishing.com. No part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means— electronic, mechanical, photocopying, recording, or otherwise—without the permission of the Darden School
Foundation. ¸
115
-2-
UVA-M-0451
The newsletter received some early publicity about its feud with the Vanguard Group over the newsletter’s original name, the Vanguard Adviser. Because the Vanguard name was the brand name of the Vanguard family of funds, they at first strongly objected to its use by Wiener as the title of his newsletter. According to their lawsuit, because Wiener had no connection with the Vanguard Group or the Vanguard family of funds, he had no right to call his newsletter the
Vanguard Adviser. The Vanguard Group dropped their lawsuit and made several concessions in
June 1993, in return for the newsletter changing its name to the Independent Adviser for
Vanguard Investors. The publicity surrounding the feud was thought to have contributed to the newsletter’s rapid growth (from 10,000 subscribers to 25,000 within a year). A description of the newsletter can be found in the promotional piece included as Exhibit 1.
Prospecting for New Customers
The newsletter relied heavily on direct-mail “prospecting” to attract new subscribers.1
Slightly over $1 million was spent in 1993 to mail promotional material to carefully selected lists of individuals in the hope that they would purchase an introductory subscription to TIAVI. The promotional package was a regular four-by-nine-inch envelope stuffed with information on the newsletter (see Exhibit 2 for the front and back views of two such envelopes). An initial oneyear membership was offered at a price deeply discounted from its usual price of $139. Sixmonth and two-year memberships were also available. The average amount received with an initial membership was $56. With a one-year membership, subscribers received 12 issues a year of TIAVI, five special reports, and one year’s access to the Fund Family Shareholder Association hotline, a private telephone number offering weekly updates and late-breaking financial news on a 24-hour-a-day basis. As the promotional material stated, “That’s a total value of $307 for only
$99, a savings of $208!”
Silver relied heavily on both MSI and Kim Scott, a free-lance circulation consultant, to select rental lists of individual names and addresses for these prospecting activities. For the most part, the selection of lists was left to the experts who knew that Silver and his partners would carefully monitor and track their results. In general, Silver thought it was a good idea to rent lists through a broker because the seller (renter) of the list paid the broker’s commission. Thus, the broker’s services were “free” to Silver when used to rent lists. In addition, the use of a broker eliminated many of the day-to-day hassles involved in prospecting—something particularly important to a small organization such as Silver’s.
Silver’s decisions, then, were not so much about which lists to use but rather about how aggressive to be. Recently, for the first time, some promotional mailings had failed to breakeven. One line of thinking within the organization was that promotional mailings should “pay for themselves.” In other words, the total cost of the mailing (printing, postage, list rental, etc.) should be recovered from the initial subscription revenues achieved from the mailing. Given that response rates almost always declined when a list was used again and again, it was argued that
1
For the purposes of this case, we will use the terms “subscriber” and “member” interchangeably. Technically, the Fund Family Shareholder Association sold memberships that included a subscription to the newsletter.
116
-3-
UVA-M-0451
once a mailing failed to break even, it made no sense to return to that list. Taken to its logical conclusion, this suggested that a promotional mailing that failed to break-even was also a mistake—the company would have been better off without it.
Exhibit 3 shows the results of the most recent promotional mailing to 10 rental lists.
Included in the last column is total net cash—calculated as subscription revenues minus prospecting costs. Several of the lists showed a loss, and the percentage of lists showing losses had been increasing over time.
Nevertheless, Silver knew he was offering a good product. About 50% of those who tried an initial subscription at the promotional rate converted it to a regular subscription at the end of the year. And the majority (75%) of subscribers renewed in each subsequent year. It cost about
$5 per subscriber to produce and send the promotional material designed to convince them to resubscribe. Because all subscription prices after the initial trial were higher, revenues were higher—about $99 on average. Fulfillment costs (costs to print, insert, and mail the newsletter) were about $12 per year.
At a recent Newsletter Publishers Association conference, Silver had obtained a copy of a formula that could be used to calculate the total number of subscriptions (circulation) to expect from a promotional mailing (see Exhibit 4). At the time, it reminded him of the dividend discount growth formula he had learned in his MBA finance class.
Silver knew he needed to respond to the growing pressure from within the company to do something about the recent prospecting losses. The $64,000 total gain from the recent prospecting effort was not nearly enough to cover his share of company fixed costs. He wondered if there was anything else he could do in addition to developing new and better promotional materials, which was something he worked on all the time. Were the losses something to worry about? Exhibit 5 presents a new set of mailing lists his brokers had recently recommended he try on his next prospecting effort. In the past, he would have mailed promotional material to all of them. But now he was not so sure.
117
-4-
UVA-M--0451
Exhibit
E
1
THE IN
NDEPENDE
ENT ADVIS
SER FOR V
VANGUARD
D INVESTO
ORS
Promotional Materrial
So ource: The Ind dependent Adviiser for Vanguard Investors. U
Used with perm mission. 118
-5-
UVA-M-0451
Exhibit 2
THE INDEPENDENT ADVISER FOR VANGUARD INVESTORS
Example Prospecting Envelopes
119
-6Exhibit 2 (continued)
Example Prospecting Envelopes
Source: The Independent Adviser for Vanguard Investors. Used with permission.
120
UVA-M-0451
-7-
UVA-M-0451
Exhibit 3
THE INDEPENDENT ADVISER FOR VANGUARD INVESTORS
Results of Promotional Mailings
List
List
CPM
$100
$145
$125
$200
$150
$100
$190
$140
$145
$160
Quantity
Mailed
49,525
47,820
39,683
45,920
40,100
41,250
35,644
48,303
38,906
37,783
In-Mail
Cost
$21,8782
$23,598
$20,072
$25,758
$21,451
$19,308
$22,067
$23,498
$20,830
$21,232
Cost per M1
$442
$493
$506
$561
$535
$468
$619
$486
$535
$562
Total
Resp.
501
321
723
340
662
578
902
128
393
526
% Resp.
AAA
BBB
CCC
DDD
FFF
GGG
HHH
JJJ
KKK
LLL
Names
Rented
50,000
50,000
50,000
50,000
50,000
50,000
50,000
50,000
50,000
50,000
1.01%
0.67%
1.82%
0.74%
1.65%
1.40%
2.53%
0.26%
1.01%
1.39%
Net
Cash
$6,1783
($5,622)
$20,416
($6,718)
$15,621
$13,060
$28,445
($16,330)
$1,178
$8,224
Total/Avg
500,000
$145.5
424,934
$219,692
$517
5,074
1.19%
$64,452
ASSUMPTIONS
Activity
Merge/Purge
Printing
Lettershop
Postage
CPM
$30
$242.50
$45
$23
Average Cash Per Order = $56
Source: Created by case writer.
1
Cost per M and CPM refer to cost per thousand.
$21,878 = 50 × ($100 + $30) + 49.525 × ($242.50 + $45 + $23).
3
$6,178 = 501 × ($56) í $21,878.
2
121
-8-
UVA-M-0451
Exhibit 4
THE INDEPENDENT ADVISER FOR VANGUARD INVESTORS
Formula for Total Circulation
Variable
Definition
Circ
Circulation: the total number of subscriptions (including conversions and renewals) resulting from a promotional mailing.
Pcs
Number of promotional pieces mailed.
resp
Response rate to the mailing (a fraction between 0 and 1).
C
Conversion rate: the fraction of responders who are later converted to a regular subscription (a fraction between 0 and 1).
R
Renewal rate: the fraction of regular subscribers who renew (a fraction between 0 and 1).
Circ
§1 C R ·
( Pcs u resp )¨
¸
© 1 R ¹
Source: Created by case writer.
122
-9-
UVA-M-0451
Exhibit 5
THE INDEPENDENT ADVISER FOR VANGUARD INVESTORS
Recommended Mailing Lists for Future Prospecting
List
QQQ
RRR
SSS
TTT
UUU
VVV
WWW
XXX
YYY
ZZZ
Type of List1
Financial
Newsletter
Health
Newsletter
Personal
Finance
Magazine
Business
Magazine
Financial
Newsletter
Political
Newsletter
Books
Tape Series
Financial
Newsletter
Financial
Newsletter
No. of Names
(thousands)
CPM
Average
Purchase2
15
$200
$70
200
$170
$35
500
$150
$30
1,000
$150
$25
75
$210
$129
250
$160
$40
200
50
$90
$90
$20
$99
30
$180
$65
40
$190
$81
Source: Created by case writer.
1
Each of these lists consisted of names of purchasers of a subscription similar to TIAVI. At the request of the company, the exact names of these subscriptions were kept confidential.
2
The average amount the individuals on the list paid to purchase the subscription.
123
124
Bibliography
"Setting Value, Not Price." Leszinski, Ralf; Marn, Michael V. In McKinsey Quarterly, Iss. 1. pp. 98−115. Published February, 1997. McKinsey & Co. Inc. (18 pages).
"Apollo Hospitals: Differentiation through Hospitality." Kulkarni, Suhruta; Makhija, Kripa;
Kumar, Unnikrishnan Dinesh. Case No. IMB425. Published 06/01/2013, Indian Institute of
Management Bangalore, (20 pages).
"Soren Chemical: Why is the New Swimming Pool Product Sinking? (Brief Case)." Rangan,
V. Kasturi; Yong, Sunru. Case No. 4188. Published 04/09/2010, Revised 11/30/2011.
Harvard Business School, (8 pages).
"Note on Behavioral Pricing." Gourville, John. Case No. 9−599−114. Published 05/25/1999,
Harvard Business School Publishing, (12 pages).
"A Practical Guide to Conjoint Analysis." Wilcox, Ronald T. Case No. UVA−M−0675.
Published 08/26/2003, Revised 11/2003. Darden Business Publishing, (9 pages).
"Fidelity Incorporated: Pricing the Fidelity Blue Chip Growth Fund." Wilcox, Ronald T. Case
No. UVA−M−0674. Published 05/20/2003, 2003. Revised 10/08/2002. Darden Business
Publishing, (5 pages).
"Portland Trail Blazers." Wilcox, Ronald T. Case No. UVA−M−0773. Published 4/3/2009,
Revised 2/22/2010. Darden Business Publishing, 2009. (11 pages).
"Customer Profitability and Lifetime Value." Ofek, Elie. Case No. 9−503−019. Published
08/07/2002, Harvard Business School Publishing, (9 pages).
"The Brita Products Co." Deighton, John. Case No. 9−500−024. Published 08/30/1999,
Revised 01/15/2002. Harvard Business School Publishing, (18 pages).
"The Independent Adviser for Vanguard Investors." Pfeifer, Phillip E. Case No.
UVA−M−0451. Published 7/19/1994, Revised 4/12/2013. Darden Business Publishing,
1994. (9 pages).
125
126
Subimal Chatterjee
Professional MBA Program (New York City)
Sept. 13 and 20, 2014
Customer Value Management: Readings and Cases
MGMT 587B: Customer Value Management
Binghamton University
THIS PRINT COURSEPACK AND ITS ELECTRONIC COUNTERPART (IF ANY) ARE INTENDED SOLELY FOR THE PERSONAL USE
OF PURCHASER. ALL OTHER USE IS STRICTLY PROHIBITED.
ISBN 1269608886
XanEdu™ publications may contain copyrighted materials of XanEdu, Inc. and/or its licensors.
The original copyright holders retain sole ownership of their materials. Copyright permissions from third parties have been granted for materials for this publication only. Further reproduction and distribution of the materials contained herein is prohibited.
WARNING: …show more content…
COPYRIGHT INFRINGEMENT IS AGAINST THE LAW AND WILL RESULT IN
PROSECUTION TO THE FULLEST EXTENT OF THE LAW.
THIS COURSE PACK CANNOT BE RESOLD, COPIED
OR OTHERWISE REPRODUCED.
XanEdu Publishing, Inc. does not exert editorial control over materials that are included in this course pack. The user hereby releases XanEdu Publishing, Inc. from any and all liability for any claims or damages, which result from any use or exposure to the materials of this course pack.
− In print copy only
− Online only
Unmarked items are available both in print and online
Customer Value Management: Readings and Cases
Table of Contents
“Setting Value, Not Price” by Leszinski, Ralf; Marn, Michael V
1
“Apollo Hospitals: Differentiation through Hospitality” by Kulkarni, Suhruta;
Makhija, Kripa; Kumar, Unnikrishnan Dinesh
19
“Soren Chemical: Why is the New Swimming Pool Product Sinking? (Brief
Case)” by Rangan, V. Kasturi; Yong, Sunru
39
“Note on Behavioral Pricing” by Gourville, John
47
“A Practical Guide to Conjoint Analysis” by Wilcox, Ronald T.
59
“Fidelity Incorporated: Pricing the Fidelity Blue Chip Growth Fund” by Wilcox,
Ronald T.
69
“Portland Trail Blazers” by Wilcox, Ronald T.
75
XanEdu Extra
An Excel−formatted spreadsheet containing the exhibits for the case above is available at http://content.xanedu.com/ds/m−0773x.xlsx “Customer Profitability and Lifetime Value” by Ofek, Elie
87
XanEdu Extra
An Excel−formatted spreadsheet containing the exhibits for the case above is available at http://content.xanedu.com/hs/503019p2.xls “The Brita Products Co.” by Deighton, John
97
XanEdu Extra
An Excel−formatted spreadsheet containing the exhibits for the case above is available at http://content.xanedu.com/hs/500024p2.xls “The Independent Adviser for Vanguard Investors” by Pfeifer, Phillip E.
115
125
Bibliography
i
ii
PRICING
Setting value, not price
© THE CURTIS PUBLISHING COMPANY
98
THE McKINSEY QUARTERLY 1997 NUMBER 1
1
279
The first task is to map benefits versus price – as the customer sees them
Bear in mind that equal value doesn’t mean equal market share
The key decision: do you stay on the line of value equivalence, or get oƒf ?
Ralf Leszinski • Michael V. Marn
A
MANUFACTURER of high-quality medical testing equipment introduces a vastly improved version of its bestselling diagnostic device at a price 5 percent higher than that of the older model it replaces. For three months, the new model is successful, gaining rave reviews from customers and increased market share. One month later, prices in the sector collapse and the company has to discount its superior new product just to maintain its traditional market share.
A highly regarded manufacturer of commercial paper prides itself on delivering extremely consistent quality and service. That consistency notwithstanding, the company is baƒfled by vacillations in its market share that accompany shiƒts from tight to loose supply in the industry.
A consumer packaged goods company executes one of the most common business tactics – it matches a competitor’s price on a large contract to supply a leading food retailer. In the months that follow, a bitter price war breaks out, destroying almost all of the industry’s profitability in this product category.
These disparate cases have at least one thing in common: apparently sound marketing strategies and tactics that produced unexpected and costly results.
But could they have been avoided? Here we will explore how these and other common and expensive marketing missteps might be averted by applying a discipline called “dynamic value management” to the pricing and product positioning that are at the core of what most marketers do.
“Value” may be one of the most overused and misused terms in marketing and pricing today. “Value pricing” is too oƒten misused as a synonym for low
Ralf Leszinski and Mike Marn are principals in McKinsey’s Atlanta and Cleveland oƒfices, respectively. Copyright © 1997 McKinsey & Company. All rights reserved.
THE McKINSEY QUARTERLY 1997 NUMBER 1
2
280
99
SETTING VALUE, NOT PRICE
price or bundled price. The real essence of value revolves around the tradeoƒf between the benefits a customer receives from a product and the price he or she pays for it.
The management of this tradeoƒf between benefits and price has long been recognized as a critical marketing mix component. Marketers implicitly address it when they talk about positioning their product vis-à-vis competitors’ oƒferings and setting the right price premium over, or discount under, them. Marketers frequently err along the two dimensions of value management, however. First, they fail to invest adequately to determine what the “static” positioning for their products on a price/benefit basis against competitors should be. Second, even when this is well understood, they ignore the “dynamic” eƒfect of their price/benefit positioning – the reactions triggered among competitors and customers, and the eƒfect on total industry profitability and on the transfer of surplus between suppliers and customers.
To illuminate the nature and magnitude of this missed value-management opportunity, value needs to be defined properly. Customers do not buy solely on low price. They buy according to customer value, that is, the diƒference between the benefits a company gives customers and the price it charges.
More precisely, customer value equals customer-perceived benefits minus customer-perceived price. So, the higher the
Exhibit 1 perceived benefit and/or the lower the price
Value map of a product, the higher the customer value
Stable market shares and the greater the likelihood that customers will choose that product. (We will return to this later.)
Perceived price
Static value management
Value equivalence line (VEL)
Many marketing and strategic assessments can be made by using a simple tool called a value map, and by considering how customers are distributed within the map for a given segment.
The value map explores the way customer value and the price/benefit tradeoƒf work in real markets for a given segment (Exhibit 1).
Customer value = perceived benefits minus perceived price
The horizontal axis quantifies benefits as perceived by the customer; the vertical axis shows perceived price. Each dot represents a competitor’s product or service.
Higher-priced, higher-benefit competitors are toward the upper right; lowerpriced, lower-benefit competitors are at the lower leƒt.
Customer-perceived benefits
100
THE McKINSEY QUARTERLY 1997 NUMBER 1
3
281
SETTING VALUE, NOT PRICE
If market shares hold constant (and if you have the right measurement of perceived benefits and perceived prices), then competitors will align in a straight diagonal line called the value equivalence line (VEL). At any desired price or benefit level, there is a clear and logical choice for customers on the VEL. So competitors aligned on the VEL say in such a market that “you get what you pay for.” The clarity of that choice almost defines a market in which shares are stable. (Note that while market shares might be stable for competitors along the VEL, their shares might not be equal. Again, more on this later.)
Exhibit 2
Share loser
E
C
A
Share gainer
VE
ad V
va
alu nt e ag e
D
L
The opposite is true for competitor E, which finds itself in a value-disadvantaged position above the VEL. Competitor E will be a share loser if the value map has been constructed properly.
Perceived price
di s ad Val va ue nt ag
e
If, however, market shares are changing, then share gainers will be positioned below the VEL in what is called a “value-advantaged” position. Competitor
A in Exhibit 2 is value-advantaged and should logically be gaining market share. If a customer is searching for a product in the benefit range of A and
B, then he or she would be more likely to choose A, since A provides the same level of benefits as B but at a lower price. Likewise, if a customer were searching for a product in the price range of A and
C, he or she would probably choose A over Value map
C, since A provides greater benefits than C
Changing market shares but at the same price. So A, positioned below the VEL that B and C reside on, oƒfers more customer value than B or C, and
B
therefore more customers prefer it.
While the marketing concepts that underpin
Customer-perceived benefits the value map are basic, advanced market research techniques (conjoint analysis, discrete choice analysis, and multi-staged conjoint analysis, for example) allow an accurate quantification of the perceived benefit dimension and its tradeoƒf against price. These advances make the eƒfective application of value maps easier than ever for marketers. That said, examples abound of costly positioning errors that could have been avoided through the use of this tool.
Illustrative case: Alpha Computer Company
The Alpha Computer Company’s experience illustrates the value map’s power, even when applied in a simple, static fashion. Alpha Computer supplied minicomputers for use primarily as servers in network applications. Alpha prided itself on its engineering skills and ability to deliver high levels of technological performance at reasonable cost. In an attempt to diagnose unexpectedly poor
THE McKINSEY QUARTERLY 1997 NUMBER 1
4
282
101
SETTING VALUE, NOT PRICE
Exhibit 3
Minicomputer value map
Alpha’s perception
Ace
market acceptance of its new line of minicomputers, Alpha created a value map that reflected its perception of the price/benefit positioning of competitors Ace Computer and Keycomp, and itself (Exhibit 3).
VE
L
Perceived price
Keycomp
Alpha believed customers chose minicomputers on the basis of two technological attributes: processor speed in MIPS (millions of instructions per second), and secondary
Alpha
access speed, that is, how quickly the computer accessed data from an external storage device such as a hard disk drive. Ace
Computer was the premium competitor: it had the highest processor speed and
Customer-perceived benefits secondary access speed, but also the highest
• Processor speed (MIPS)
• Secondary access speed price. Keycomp not only had slower processor speed and secondary access speed than Alpha but was also priced 10 to 15 percent higher. So, Alpha thought that Keycomp was value-disadvantaged and that Alpha itself was valueadvantaged.
If Alpha’s perception of the value map in Exhibit 3 were correct, then Alpha should have been gaining market share and Keycomp losing it. The opposite was occurring, however, and Alpha’s managers were baƒfled. They thought their product was superior to Keycomp’s at a lower price, and they could not understand why it was not a huge success.
Alpha’s problem was a common one. It did not understand the customerperceived attributes that really drove customer choice of minicomputers.
Alpha’s marketing department commissioned research to try to confirm its hypothesis that processor speed and secondary access speed were indeed the most important features. Sixty buyers were questioned about their criteria for selecting a network minicomputer supplier.
Much to Alpha’s surprise, processor speed and secondary access speed ranked only fourth and sixth on their list. Soƒtware and hardware compatibility, perceived reliability, and quality of vendor technical support ranked above raw processor speed. Even quality of user documents (the manual that accompanies the hardware) ranked above secondary access speed.
As it turned out, processor speed was indeed important, but most customers had a minimum processor speed requirement that all competitors easily exceeded. However, the nature of most network applications made secondary access not that important. In fact, Alpha was understood
102
THE McKINSEY QUARTERLY 1997 NUMBER 1
5
283
SETTING VALUE, NOT PRICE
by customers to be slightly better than Keycomp on processor speed and secondary access speed, but these features just did not matter that much to them.
The research also showed that Keycomp was highly rated on compatibility, reliability, vendor support, and user documents. Alpha, on the other hand, fell short on these. Its operating system soƒtware and hardware plug configuration created compatibility problems for many customers. Some remembered reliability problems with an earlier generation of Alpha’s minicomputer that tainted their perception of its new product. Alpha’s technical support Minicomputer value map was considered diƒficult to get hold of and its
Actual customer perception user documents were seen as the weakest in the industry.
Ace
Perceived price
Keycomp
Alpha
VE
L
Exhibit 4 shows how the value map was redrawn to reflect customers’ perceptions of benefits and performance rather than
Alpha’s. It showed that Keycomp performed so well on the attributes most important to customers that, despite its higher price, it was value-advantaged and therefore justifiably gaining market share. Conversely,
Alpha performed so poorly on attributes most essential to customers that, despite its low price, it was still value-disadvantaged and predictably losing share.
Customer-perceived benefits
• Compatibility
• Reliability
• Vendor support
• Processor speed (MIPS)
• Documentation
• Secondary access speed
The insights from this properly constructed value map prescribed a clear course for
Alpha. It mounted a crash program to correct the important attributes on which customers had rated it so poorly. A minor rewrite of operating system soƒtware and a simple redesign of the hardware plug configuration fixed the compatibility issue. The company then mounted an aggressive market information campaign to demonstrate the improved reliability of its latest model. Additional service representatives and toll-free access lines were put in place to enhance technical support, and user documents were redraƒted.
The results are shown in Exhibit 5. In only six months, Alpha increased customer-perceived benefits so much that it was able to increase its price by
8 percent and still gain its fair market share. The price and volume increase more than doubled Alpha’s operating profits.
The Alpha Computer case illustrates several important points about value management: THE McKINSEY QUARTERLY 1997 NUMBER 1
6
284
Exhibit 4
103
SETTING VALUE, NOT PRICE
Exhibit 5
Minicomputer value map
Repositioning Alpha
Ace
Alpha
VE
L
Perceived price
Keycomp
Customer-perceived benefits
• The key to success oƒten resides in gaining a clear understanding of the real attributes driving customer choice and their relative importance. • “Soƒter,” nontechnical attributes (perceived reliability, quality of vendor support, ease of doing business) are oƒten as important as or more important than precisely measurable technical features.
• Trusting internal perceptions of which attributes drive customer choice can be a fatal mistake; rely on customers for this critical information.
• Compatibility
• Reliability
• Vendor support
• Processor speed (MIPS)
• Documentation
• Secondary access speed
The case also shows the opportunities value maps oƒfer value-disadvantaged companies to understand their markets better. Another case, that of car maker Mazda’s experience with its Miata sports model, demonstrates the kind of opportunity that a value-advantaged company can easily forgo if it does not fully appreciate its position (see the boxed insert, “US economy sports car market, 1990”).
Distribution of customers on the value map
VE
L
Perceived price
In discussing the stability associated with a position on the VEL, and the eƒfects of competitive moves away from it, we have implicitly assumed that all positions along the line are equally attractive. This is not the case. Even for a well-defined segment, customers are not
Exhibit 6 spread evenly along the line; if they were,
Customer volume distribution every competitor on the VEL could be expected to have the same market share.
Sometimes this can be explained by hisCustomer volume torical reasons; mostly, however, it is due to the distribution of customers along the VEL
(Exhibit 6).
Customer-perceived benefits
104
History plays an important role: how long a competitor has held its position with customers oƒten explains large market share diƒferences among companies with otherwise the same value proposition. This phenomenon, also called “order of entry,” can be seen in its extreme form in deregulated utilities.
THE McKINSEY QUARTERLY 1997 NUMBER 1
7
285
US ECONOMY SPORTS CAR MARKET, 1990
I
ntroduced to the US market in 1990 at a manufacturer’s suggested retail price of
$13,800, the Mazda Miata was a retro-sports roadster that captured the imaginations of ageing baby boomer car buffs who originally fell in love with the classic British roadsters of the 1960s and 1970s made by M G and
Triumph. As much fun as its British predecessors but better built and more reliable, the Miata was an instant hit in the
United States.
Mazda Miata
Manufacturer’s suggested retail price
$ thousand
20
Mazda RX-7
18
Toyota MR-2
16
Honda Prelude
Mazda underestimated the appeal and the high perceived benefits of the simple but unique Miata. The price was disproportionately low for the perceived benefit. Mazda dealers, however, recognized this price/benefit imbalance and claimed the surplus for themselves in the form of $2,000–3,000
“ market price adjustments” that they added to the suggested retail price (and which customers gladly paid).
14
Nissan 240SX
Mazda Miata
12
0
Perceived benefit
A new competitor oƒfering similar or even slightly better value than an incumbent telephone or electricity company will not provoke the significant changes in consumer purchasing that might be expected.
A more important and probably more common explanation of market share diƒferences among competitors on the VEL is the distribution of customers along this line. Typically they are not distributed evenly, but clustered. There are several reasons for this. Sometimes consumers are not equally aware of the true nature and availability of competing products. Companies might use different channels to reach consumers, or their salesforces might not adequately communicate benefits to customers. If so, a gap can exist between customers’ perceptions of a product’s benefits and the benefits that it actually delivers.
Even in a perfect world, consumers would be unevenly distributed along the VEL because they do not necessarily view benefits and prices in a linear way. There are benefit-bracketed customers who explicitly want minimum or maximum benefit levels and find positions on either side unacceptable.
Market research shows that break-points exist for some products and services at which a small increase in the benefits oƒfered will lead to a large increase in the value a customer perceives. Some buyers of automotive components, for example, will not accept delivery reliability below a minimum level. Some computer buyers, on the other hand, do not value additional memory beyond a certain level because existing memory more than satisfies their needs.
THE McKINSEY QUARTERLY 1997 NUMBER 1
8
286
105
SETTING VALUE, NOT PRICE
A second group is price-capped customers who are unwilling to spend more than a fixed amount for a particular product or service. The price of the average home PC has held at about $2,000 for several years, even though performance has improved sharply. This could indicate that there are pricecapped customers at around this level who are unwilling to spend more even if they could get more features. Only customers who fall into neither category, benefit-bracketed or price-capped, are actually willing to consider the full range of tradeoƒfs along the VEL.
Understanding volume distribution along the VEL is therefore crucial to making an intelligent decision about product position. In many cases, however, it is poorly understood, leading to wrong decisions. Typical mistakes are:
• Positioning an apparently competitive product at a low-volume part of the VEL and not getting the expected volume gains. A maker of metalcoating machinery positioned a new product technically half way between two competing products, hoping to pull in customers not entirely satisfied with these. What it had not realized was that there was no significant volume between the two extremes, because each answered a specific speed requirement of downstream customers. Failing to understand that there was no demand for a medium-speed machine, even one that was competitive on technical specification and price, forced the manufacturer to take a multimillion-dollar writeoƒf.
• Positioning a product too high or too low on the VEL, thereby inadvertently excluding a large portion of price-capped or benefit-bracketed customers.
The drastic fall in demand for one company’s supercomputers is an example of this. Even though the company’s ever more powerful machines remained on the VEL, there was no longer a customer imperative for all that processing power to be concentrated in one machine, as more broadly distributed processing had become preferred by most users.
Dynamic value management
Alpha Computer and Mazda Miata illustrate the pitfalls of failing to understand the “static” value positioning of a product or service. But getting a product to the right position on a static value map is only part of managing value eƒfectively. Unfortunately, neither competitors’ positions on a value map nor customers’ perception of products and suppliers are frozen in time.
Value maps are not static but dynamic, constantly changing in important and oƒten predictable ways.
Any change in product positioning by one competitor, be it cutting price or improving features, will lead others to move, either to preempt shiƒts in market share or to react to them. We apply the term “dynamic value manage-
106
THE McKINSEY QUARTERLY 1997 NUMBER 1
9
287
SETTING VALUE, NOT PRICE
ment” to the discipline of managing price/benefit positioning not just in a static fashion, but with explicit and thoughtful consideration of likely changes in competitive value positions and customer value perception.
Companies that master this discipline can reap huge rewards and avoid equally huge pitfalls.
Illustrative case: MTE
Perceived price
MTE is the manufacturer of high-quality medical testing equipment mentioned at the beginning of this article. Its primary product was a blood diagnostic testing machine used in highExhibit 7 volume hospital laboratory applications. Dynamic value management
MTE was the recognized premium supplier
Leader enhances benefits but does not raise
(with the highest price and benefits) in a price commensurately… stable market that included three other leading competitors (Jackson, PZJTech, and
Labco) positioned squarely on the VEL
MTE
(Exhibit 7).
Jackson
VE
L
As is oƒten the case, MTE, as the premium
PZJTech
supplier, was the real innovator in this market. The improved version of its blood
Labco
diagnostic testing machine was more accurate and had faster testing cycle times.
But MTE was in a dilemma over how to price its terrific new model. Research showed
Customer-perceived benefits the added benefits would justify a 10 percent price increase and still keep the model on the
VEL – that is, MTE would hold its market share. But, equally, it could keep the price the same and position the new model in a highly value-advantaged position in the hope of gaining significant market share.
MTE decided on a compromise, raising its price by 5 percent, a meaningful increase that still kept it in a value-advantaged position (the dotted circle in
Exhibit 7). The response was instant and positive. Customers recognized the
5 percent increase was a small premium to pay for enhanced accuracy and cycle times. The machine sold well and immediately increased MTE’s share of the market.
This success, of course, was at the expense of Jackson, PZJTech, and Labco, none of which had the expertise or resources to introduce products to rival
MTE’s new model. Faced with falling sales, they took the only measure they could to defend their market shares – they lowered their prices by at least
5 percent (Exhibit 8). The market shares of all four companies quickly returned to their previous levels, but at the lower prices. As Exhibit 8 shows, the VEL had simply shiƒted downward and MTE’s value-advantaged position was
THE McKINSEY QUARTERLY 1997 NUMBER 1
10
288
107
SETTING VALUE, NOT PRICE
Exhibit 8
Dynamic value management
O ld VE
L
…resulting in a downward price shift across market
VE
L
Jackson
N ew Perceived price
MTE
essentially nullified. The lowered
VEL was good for customers because they got more for their money, but the suppliers got less for their products. It represented a wholesale transfer of market surplus from suppliers to customers.
Could MTE have managed the value dynamics of this situation better?
Possibly. If it had raised the price of
Labco
its new model by 10 percent and positioned it on the existing VEL, it would have held its traditional share but at a 5percenthigherprice.Jackson,
Customer-perceived benefits
PZJTech, and Labco, experiencing no loss of market share, would probably not have reacted at all. Industry prices would have been maintained, and
MTE’s profit would have risen significantly.
PZJTech
Changing your position in a dynamic world
Marketing managers have two basic options for improving their products’ position, regardless of whether they are in a proactive or reactive situation.
They can reposition their product along the VEL, or move oƒf it. These diƒferent moves engender very diƒferent outcomes – diƒferent competitor and customer reactions and diƒferent prices, volumes, profits, and risks.
Repositioning along the VEL
Repositioning a product along the VEL, usually a less aggressive move, requires a company to understand where customer clusters are on it, and how other competitors are positioned in relation to them. The decision of whether and how far to move should include the following steps:
Understanding and weighing the risks and opportunities. Repositioning a product is likely to lose some customers who preferred the old positioning.
Equally, it will gain customers who prefer the new positioning. Failure to understand this tradeoƒf could lead a company to surrender a good customer franchise in exchange for a reduced, and probably more competitive, new franchise. Being smart about choosing the right attributes to vary. Customers do not consider all product attributes to be equally important; there is therefore more “bang for the buck” in changing some attributes rather than others.
The knack is to select the features that will attract new customers without
108
THE McKINSEY QUARTERLY 1997 NUMBER 1
11
289
SETTING VALUE, NOT PRICE
losing old ones, that have the greatest impact on customers, and that the company can provide cost-eƒfectively.
Knowing what price change is appropriate for a given attribute change.
If the aim is to stay on the VEL, any change in benefits must be accompanied by a price change. Not increasing the price enough will force competitors to match the new positioning, leading to an unwanted industry price decline
(as with MTE); raising the price too high will lead to a volume loss. Market research tools such as conjoint analysis can determine the magnitude of change required.
Choosing those changes least likely to provoke undesirable competitive reactions. If the repositioning is successful, or looks as if it will be, competitors will react. The likeliest, and least desirable, reaction is a price cut, which oƒten leads to price cuts across the industry and lower profits for all. One manufacturer of medical supplies always reacted to competitors’ price cuts by improving benefits. Every time a competitor dropped its price, the supplier countered with an improved version of its product at the same price, but on the new VEL. In this way it gained a distinctive market position, offering increasingly superior benefits over competitors that chose to move only along the price dimension.
Choosing the new position along the VEL. There are two options: either to move to a new position within the extremes defined by current competitors, or to move to a new position beyond the current extremes. There are diƒferences in risk and potential competitive moves between the two:
• The success of a new positioning within current competitive extremes depends on locating the right customer concentration and standing out from competitors. As this approach seldom expands a market, competitors will probably react to their declining sales.
• Moving to a new position along the VEL outside the existing extremes can expand a market. While the upside opportunities can be greater (and the threat of retaliation lower), success depends on a thorough understanding of the size and needs of the latent demand that the new product or service is designed to meet.
Moving oƒf the VEL
A move oƒf the VEL into value-advantaged territory might seem attractive on the surface. As the experiences of many companies show, however, such a move requires an even better understanding of the dynamics, risks, and opportunities than do moves along the VEL.
What is diƒferent about moves oƒf the VEL? A repositioning along the VEL is likely to threaten only one or two neighboring competitors currently on the
THE McKINSEY QUARTERLY 1997 NUMBER 1
12
290
109
SETTING VALUE, NOT PRICE
Perceived price
line. Moving below the VEL oƒten threatens all competitors, because such moves usually define new and lowered VELs that force them to reconsider their own positions. Only rarely does the
Exhibit 9
VEL move upward; to do so would require
Repositioning off the value equivalence line customers to accept the actual value reduction and most suppliers to move in the
Expanding the horizon same direction.
Addressable customer horizon
A
VE
L
A
When a product is repositioned below the
VEL, its “horizon” of potential customers grows (Exhibit 9). Take, for example, an electric drill whose power was increased but which was sold for the same price. The new product appeals not only to customers who initially bought it, but also to those who had previously paid more for a drill with the higher power rating.
Just moving oƒf the VEL to expand the horizon of customers does not guarantee success, however. Market research must first establish that the expanded horizon does indeed include new concentrations of customers, not just empty space.
Customer-perceived benefits
Likely competitive reactions to moves oƒf the VEL
In today’s highly competitive markets, rivals seldom passively accept volume or market share losses. They usually react by trying to improve their products by selectively adjusting attributes, or by dropping price. How they will react is a function of a number of parameters, including:
The type of change that set the whole process in motion. The typical reaction to a competitor’s move is to try to counter along a similar axis. If the salesforce reports massive price cuts by a competitor, they will want to reciprocate. If a competitor introduces a new service, the salesforce will want to oƒfer something similar. A first mover’s repositioning along the benefits axis tends to damage profits less than price reductions would. It is also easier to retract benefits that are rejected by the market or are uneconomic to provide, than to try to raise prices aƒter a round of reductions.
Competitors’ strategic mindset. The degree of volume and profit pressure a competitor is under and its understanding of the economics of price changes (for example, how price and volume trade oƒf against profit) will drive the type of reaction it makes.
Even in commodity-like industries, there are examples of manufacturers successfully improving their products and services rather than cutting prices.
110
THE McKINSEY QUARTERLY 1997 NUMBER 1
13
291
SETTING VALUE, NOT PRICE
In a US specialty chemical segment, for example, the two leading companies have about 40 percent of the market. They and their customers recognize that there are no real technical diƒferences between the two suppliers’ products. So when one competitor increases its support services, the other improves its services too. While the industry is competitive, and the level of service high and rising, prices have also risen and profits have remained strong. In the past five years, neither leader has reacted to a competitor by reducing its price – a move that would surely have made the industry less profitable. Eƒfects on demand and volume distribution
Competitors’ behavior can actually shiƒt the distribution of demand along the VEL (Exhibit 10). As the line is shiƒted downward through improved combinations of price and benefit, it is not automatic that the “old” pattern of customer distribution follows suit. Some customers might be benefitbracketed, others might use the changes to rethink their own price/benefit tradeoƒfs, and, finally, new oƒfers could stimulate latent demand.
Repositioning off the value
Exhibit 10
equivalence line
ld
N
ew
VE
L
3
O
2
VE
L
“Old” and “new” customer volume distribution
Perceived price
If the distribution of demand changes, a shiƒt oƒf the VEL will not always bring the desired volume increase. The established manufacturers in one consumer durable industry assumed most customers were price-capped, and therefore had not oƒfered increased benefits. But when a new competitor introduced a new product at a significantly higher price, 30 percent of the volume shiƒted to that new product. Some consumers had been looking for more benefits aƒter all.
1
Customer-perceived benefits
A move oƒf the VEL has to be large enough for customers to notice and attractive enough to make them want to try the repositioned product. Marginal moves oƒten backfire. If consumers do not perceive enough diƒference to make them switch supplier, but competitors, which follow such moves closely, decide to copy it, the VEL can quickly drop without aƒfecting market shares, but lowering price and profit.
In the case of a company that installed heating equipment, the information that its key competitor had cut the cost of installation labor by 5 percent led it to cut its own price too. Unfortunately, this company did not adequately consider the basis on which architects and contractors compare bids – that is, the total installed costs. The selective 5 percent drop in labor reduced the
THE McKINSEY QUARTERLY 1997 NUMBER 1
14
292
111
SEGMENTATION THEORY
I
n classic marketing and segmentation theory, a segment is defined as a group of customers with identical needs and buying behavior. In theory, all customers constituting a precisely defined segment would be equally receptive to all products positioned anywhere along the
VEL. Therefore, all products positioned on the
VEL should have the same market share.
In practice, this is clearly not the case.
There are two ways to resolve the conflict:
• Define a segment (and the products positioned in it) so that it contains all the realistic and feasible alternatives customers consider for a given purchase, and accept that there will be some differences in buying decisions. We will take the second approach, as it makes the concepts in this article easier to apply in the messy “real” world, without compromising the quality of the answer.
• Define each segment so narrowly that it contains only one customer.
total installation cost by less than 1 percent – too slight a diƒference for the market to notice.
Moving oƒf the VEL therefore requires two decisions about the direction and the distance:
• Direction. What are the customer volume elasticities of moves along the price axis and the benefit axis (by attributes)? Do I want to increase my benefits, lower my price, or both?
• Distance. How far do I have to move from the VEL to expand my horizon of customers suƒficiently? How far do I have to move to diƒferentiate myself from competitors in the eyes of a group of potential customers? How strong will competitors’ reactions be? How many additional benefits can I aƒford to deliver and what price cut am I willing or able to absorb?
Moving below the VEL is always a risky strategy that can, if executed well, reap some benefits. In many cases, however, too little thought is given to what customers actually want, how competitors will react, and how demand might change as a result of competitors’ moves. This negligence can lead to profit declines where once there were high hopes.
Using dynamic value management to respond to external changes Dynamic value management can also be a powerful tool to help prescribe reactions to changes in competitive position or customer needs. A competitor’s actions can set in motion the same set of dynamics. Dynamic value management is as useful in determining reactions to such moves as it is in initiating them.
Competitors’ moves
Being on the receiving end of a competitive move demands an approach similar to the proactive stance above. It also requires a cool head. If the
112
THE McKINSEY QUARTERLY 1997 NUMBER 1
15
293
SETTING VALUE, NOT PRICE
salesforce is sending panicky messages about competitive price cuts, pressure is created to act quickly. In most cases, the easiest lever to pull in the short term is price. And in all too many cases, this would be a mistake. A series of thoughtful decisions using the dynamic value-management approach can help formulate a more eƒfective and less costly response. A set of questions should be answered:
• Do customers perceive the competitor’s move as a move oƒf the VEL?
To find out, ask the customer. Too oƒten this question is answered hastily and wrongly on the basis of hearsay from the field. If the move is not perceived to be a wholesale jump to a new VEL, there may be no need to react.
• If the competitor has moved oƒf the VEL, has its “horizon” expanded suƒficiently to draw in new customers? If market research shows it has not, again there is no need to react.
• If new customers are buying the competitor’s oƒfering, are they our customers or somebody else’s? The answer to this question determines not the need for a reaction, but the speed and extent of it. If the primary threat is to somebody else’s customers, let them react. All competitors will be likely to react eventually, but timing is important. A gradual cascade of reactions not only will prevent panicky overreactions, but can also create opportunities to observe informative customer buying behavior.
• If a reaction is needed, how strong should it be? Should it be a surgical strike on one product, channel, or market, or across the board? Should it entail price changes, benefit changes, or a combination of both?
Cyclical markets
In cyclical industries, the value map can change not only because of competitors’ moves, but also because of changes in customer needs over the course of the cycle. The following case illustrates the enhanced challenge of dynamic value management in highly cyclical businesses.
Illustrative case: Pace Paper Company
The Pace Paper Company produced high-grade paper for business forms, brochures, and corporate annual reports. Pace and its two main competitors,
Marco Paper and Valentine Paper, sold directly to large regional and national printing companies. Demand for this high-grade paper tended to vary wildly with the overall economic cycle.
Pace produced paper of unsurpassed quality and consistency and provided equally consistent delivery service. But it found itself gaining market share in down markets when there was excess supply and losing it sharply in times of tight supply. Given its consistency and quality throughout the
THE McKINSEY QUARTERLY 1997 NUMBER 1
16
294
113
SETTING VALUE, NOT PRICE
economic cycle, Pace could not understand the drastic market share shiƒts it regularly experienced.
The problem was that while Pace stayed the same, customers changed through the cycle. Exhibit 11 shows the value map for this market at diƒferent stages of the cycle. At the bottom of the cycle (when supplies were plentiful),
Exhibit 11
Value position changes through a cycle
Industry supply /demand balance
Excess supply
Tight supply
Stable market shares
Benefit attributes ranked
1. Paper quality/consistency
2. Order lead time
3. Order fill rate
Pace
Marco
Valentine
Pace
Marco
Benefit attributes ranked
1. Order lead time
2. Order fill rate
3. Paper quality/consistency
Valentine
Time
customers had no problem obtaining enough paper. They therefore demanded high and consistent quality so that printing jobs would run eƒficiently through their plants with minimum rejects. Order lead times (the number of days between an order being placed and the paper being delivered) and order fill rates (the percentage of the total order carried on the first shipment) did not much matter, since printers usually had ample safety stocks of paper in their own warehouses. Despite being slightly higher priced than Marco and Valentine, Pace’s paper quality and consistency were so superior that it was value-advantaged in times of excess supply and gained market share – as shown at the lower leƒt of the value map in Exhibit 11.
As supplies tightened, however, printers oƒten found their stocks depleted.
They became increasingly concerned about running out and having to shut down their printing plants temporarily. They therefore relaxed quality and consistency requirements in favor of delivery performance. As the value map at the upper right of Exhibit 11 shows, paper quality and consistency slipped to third place behind order lead time and order fill rate. Valentine Paper could not match Pace’s quality and consistency, but its order lead times and fill rates were better than Pace’s. The result was that in times of tight supply,
Valentine would shiƒt to a position of value advantage (and thus gain market
114
THE McKINSEY QUARTERLY 1997 NUMBER 1
17
295
SETTING VALUE, NOT PRICE
share), while Pace would slip to a value-disadvantaged position (and, of course, lose share).
Armed with the insights provided by the value maps in Exhibit 11, Pace’s managers embarked on a project to determine how they might improve their order lead times and fill rates in times of tight supply. They discovered that if they relaxed their product consistency slightly (in a way that was almost imperceptible to customers), they could increase plant throughput enough to cut order lead times, increase order fill rates, and return to value equivalence in tight markets. When supplies loosened, Pace reverted to its original level of paper consistency to reinforce its traditional valueadvantaged position. This fine-tuning over the course of market cycles enabled Pace to maintain its share in tight markets without cutting prices or jeopardizing future positioning in down markets.
Looking ahead
With product life cycles shrinking (measured in months rather than years in the computer industry, for instance), customers becoming more sophisticated and demanding, and tougher local and even global competitors emerging in most markets, value maps are shiƒting at faster rates than ever. Fortunately, advances in market research techniques make the execution of eƒfective dynamic value management easier than ever.
The discipline of dynamic value management not only promotes sustainably improved market performance and profitability, but also yields a number of attractive side benefits, including:
• More genuine closeness to customers, thanks to a richer, more externally driven understanding of the benefit attributes that really matter to customers
• An enhanced understanding of competitors: their strengths in the eyes of customers, their strategies, and their likely reactions to price and benefit moves by your company
• More integrated product/market strategy formulation, where the linkages between price, benefit delivery to customers, competitor capabilities, and changing customer preferences are explicit.
The payoƒf for getting dynamic value management right has probably never been higher; the consequences of getting it wrong, never more devastating.
For a growing number of companies, dynamic value management is providing a compass for navigating the increasingly unstable seas of change and uncertainty that challenge most marketers today.
THE McKINSEY QUARTERLY 1997 NUMBER 1
18
296
115
IMB 425
SUHRUTA KULKARNI, KRIPA MAKHIJA AND U DINESH KUMAR
APOLLO HOSPITALS: DIFFERENTIATION THROUGH HOSPITALITY
The ‘‘wow’’ factor in service relies on constant innovation and demands continuous and sensitive focus on all issues that may affect the patient’s stay in a hospital. Every touch point of the hospital needs to be ‘‘alive’’ and the client must be able to feel the warmth offered. The culture of service is imperative in today’s scenario, where the differentiators could just be the manner in which services are offered. All the major players could replicate infrastructure within a short span of time, but not the service culture.
Dr. Umapathy Panyala, Chief Executive Officer, Apollo Hospitals, Bangalore (March 2013)
Dr. Panyala, CEO, Apollo Hospitals, Bangalore believed that in the future, the hospitality aspect of hospitals—the service provided to patients—would differentiate Apollo Hospitals from a large number of equally competent competitors in the growing Indian healthcare industry. He had set up a quality department at the Apollo Hospital in
Bangalore, headed by Dr. Ananth Rao. Apart from being an expert on Metabolic Diseases and Biochemistry, Dr.
Rao was also a Lean Six-Sigma black belt from the Indian Statistical Institute, Chennai.
You can’t manage what you don’t measure—although this may sound clichéd; I am still a firm believer of this philosophy and want to apply this, especially in the hospitality part of hospitals.
Clinical benchmarking is a compulsory requirement and is taken care of; however, patients have so many other touch points in their stay at hospitals—the hospitality part. Some of the world-class hotels (such as the Ritz–Carlton) have performed benchmarking to standardise their hospitality offerings; at the same time, its employees are allowed to use their judgment to provide whatever delights the customer in every visit. 1 We want to internalise this in our culture as well.
– Dr. Ananth Rao, Head–Quality Department, Apollo Hospital, Bangalore (March 2013)
Dr. Rao also believed that the hospitality aspect would differentiate Apollo Hospitals from its competitors. Patient cure and care played very important roles in hospitals. Many patients were generally anxious when in a hospital and the sense of disservice increased their anxiety level very easily. Integrating healthcare and hospitality was essential for creating patient-focused care. Hospitality aspects included a smooth admission procedure, friendly medical and non-medical staff, comfortable furniture, varied choices on the food menu, attractive surroundings, recreational facilities, privacy, clear signposting, adequate provisions for visitors, and so on. 2 Important aspects of hospitality were managed by the nursing staff and other non-medical staff, which meant inherent variability of service owingto human interventions.
Dr. Panyala and Dr. Rao wanted to measure the hospitality aspects at Apollo Hospitals and improve hospitality to create a world-class hospital. Dr. Rao and his team collected feedback every day from the patients and received a number of complaints, ranging from not having a TV remote to long response time on the part of nursing staff in attending to requests from patients. For Apollo Hospitals, it was important that the patients’ experience in the hospital was not compromised, since it could have a significant financial impact. Managing the hospitality elements of the hospital was as important as managing the clinical aspects. Apollo Hospitals had a stringent process in place to take care of clinical aspects. Dr. Rao wanted to improve the hospitality at Apollo Hospitals by reducing the
1
Hall, J. M. and Johnson, M. E., When should a process be art, not science, Harvard Business Review, 2009, 1–9.
Hepple, J.,Kipps, M. and Thomson, J., The concept of hospitality and an evaluation of its applicability to the experience of hospital patients,
International Journal of Hospitality Management, 1990, 9(4),305–318.
2
Suhruta Kulkarni, Kripa Makhija and U Dinesh Kumar, Professor of Quantitative Methods and Information Systems, prepared this case for classroom discussion. V Sandeep assisted in data collection and analysis. This case is not intended to serve as an endorsement or source of primary data, or to show effective or inefficient handling of decision or business processes.
Copyright © 2013 by the Indian Institute of Management Bangalore. No part of the publication may be reproduced or transmitted in any form or by any means – electronic, mechanical, photocopying, recording, or otherwise (including internet) – without the permission of Indian Institute of
Management Bangalore.
19
Page 2 of 20
Apollo Hospitals: Differentiation through Hospitality
number of complaints from patients; he also wanted to achieve significant improvement in sigma levels measured through the Six Sigma performance scale. According to Dr. Ananth Rao:
The immediate challenge is to understand the patients’ sentiment towards the hospitality provided and to design a process improvement plan that is affordable. Apollo takes feedback from patients every day and the quality department staff interviews many patients every week to understand their needs.
Dr. Rao was aware that improving hospitality at Apollo Hospitals was going to be a continuous exercise in improvement; collecting feedback was one way of approaching the process of continuous improvement. He treated every complaint as a “defect” and planned to use lean Six Sigma concepts to eliminate defects. Implementing Six
Sigma in all departments was likely to be a challenge since departments such as housekeeping faced high attrition rates. His immediate objective was to introduce a system where future complaints related to hospitality could be reduced.Also, how much importance should be given to hospitality by Apollo Hospital was one of the dilemmas faced by Dr Rao and he wanted to set a realistic target for Sigma level in hospitality at Apollo.
APOLLO HOSPITALS: THE TRENDSETTER
Dr. Prathap C. Reddy, founder of Apollo Hospital Enterprises Ltd. (AHEL) had accomplished a successful medical career in the United States. He returned to India in 1972 to contribute to the healthcare system in India. Health infrastructure in India was very poorly developed in the 1970s. In 1971, there were 3,862 hospitals and 12,180 dispensaries with a total of 348,6553 beds for a population of 548,159,6524—a ratio of 6.36 beds per 10,000 people as against the ratio of 9 beds per 10,000 people in 2011. 5 India’s first National Health Policy was declared in 1983,6 almost 36 years after independence, which was an indication of the neglect faced by the health sector in the country since independence.
Dr. Reddy had set up a good medical practice in India and used to send patients outside the country for specific treatments. However, in 1979, a young patient died as he could not arrange the money for treatment in the United
States. Dr. Reddy then decided to provide the best of medical treatment from the West to patients in India with an emotional touch, calling it “High Tech with High Touch.” Apollo was a doctor-promoted enterprise—10,000 Indian doctors, 4,700 U.S.-based doctors, and 60 doctors from the United Kingdom invested approximately USD 5,000 to start the venture. Dr. Reddy selected the best of the talent available to ensure the best possible service and care. He also ensured that a clear distinction was maintained between business management and clinical management. 7Apollo pioneered world-class healthcare in India, which was later emulated by several other hospitals.
Apollo focused on technological excellence and garnered many firsts to its credit in the country. Apollo was the first not only in India but also in South Asia to launch Oncological Robotic Surgery, G4 Cyberknife Robotic
Radiosurgery System, 320-slice computed tomography scanner, 64-slice positron emission tomography-computed scan system, full-field digital mammography with tomosynthesis, and many such technologies. 8 According to Dr.
Rao, Apollo intended to carry forward technological excellence in hospitality to provide patients with the best cure and care services.
Dr Preetha Reddy, Managing Director, Apollo Hospitals Enterprises Limited has been the pioneer and chief architect of the tender loving care –TLC ‘‘mantra’’, a pillar of the Apollo way, which is affectionately applied to every patient at Apollo Hospitals. “The patients and staff comprehend this language better,” she points out. The concept of TLC integrates service delivery with clinical outcomes resulting in exceptional patient experiences9,10.
3
Background Papers: Financing and Delivery of Healthcare Services in India, National Commission on Macroeconomics and Health, Ministry of
Health and Family Welfare Government of India, 2005, p. 47.
4
Source: http://cyberjournalist.org.in/census/cenpop.html, accessed on March11, 2013.
5
Source: http://www.globalhealthfacts.org/data/topic/map.aspx?ind=78, accessed on March11, 2013.
6
Health Research Policy, Indian Council of Medical Research, New Delhi, (October 2007).
7
Mitra, M., The Apollo Mission, Corporate Dossier with The Economic Times, June 1, 2012.
8
Apollo Investor Presentation, www.apollohospitals.com, accessed in January 2013.
9
N Amarnath, and D Ghosh, The Voyage to Excellence: The Ascent of 21 Women Leaders of India Inc., Pustak Mahal, pp. 80-95.
10
http://www.apollohospitals.com/apollo_pdf/dr_preetha_reddy_managing_director.pdf
20
Page 3 of 20
Apollo Hospitals: Differentiation through Hospitality
CLINICAL BENCHMARKING
Apollo Hospitals had been using a clinical score card called ACE@25 (Apollo Clinical Excellence), which measured and monitored clinical excellence among the group’s hospitals. ACE@25 measured 25 clinical parameters
(Exhibit 1) every month, which were benchmarked against global standards. ACE@25 was launched on
September18, 2008 and used across 32 hospitals of the group. Clinical benchmarks were published by various institutions and bodies such as Cleveland Clinic, Mayo Clinic, and National Healthcare Safety Network (NHSN), among others. Hospitals were grouped according to their bed strengths, locations, services offered, and so on. Group
A hospitals had to report 25 parameters—23 were common parameters, while two were location-specific. Group B and Group C hospitals had to report 15 and 10 parameters, respectively, out of which two were location-specific.
ACE@25 was an internally developed clinical scorecard, created by drawing upon the wealth of expertise available within Apollo. An audit committee at the corporate level was set up to validate the data, methodology, and definitions followed at each location. According to Sangita Reddy, Executive Director, Apollo Hospitals Group:
We needed a yardstick like ACE@25 that would pit us against international institutes like
Cleveland Clinic, Mayo Clinic, and others, and position us on the global healthcare firmament for excellence in clinical quality. This also enables us to assess where we stand and where we need to
11
be, while pursuing excellence in clinical quality.
Apart from this internal benchmarking exercise, seven of Apollo’s hospitals were accredited by the Joint
Commission International (JCI); and it was the largest group in South Asia to be accredited by the JCI. The JCI was a U.S.-based accreditation body dedicated to improving healthcare quality and safety around the world and recognized as the gold standard for hospitals. Apollo was also accorded the Superbrand status by the Indian
Consumer Superbrands Council, which recognised that the best practices were used in the brand. Apollo was the
12
only hospital that was accorded the Superbrand status in India. There were other accreditations that several Apollo hospitals had achieved (Exhibit 2).
According to Dr. Panyala,
Living the brand should be our focus in every initiative or activity we perform. Apollo Hospitals has been one of the consistent names among the Superbrands. The perceived value of a brand like
Apollo Hospitals is set very high in the backdrop of the decades of service and excellence it has offered. Clients need to see and experience that value, and the gap between perceived value and obtained value must be zero at best or at a bare minimum.
PATIENTS’ FEEDBACK AND REAL-TIME ACTION
On average, a patient spends 80% of the time in hospital for the care part rather than the cure, and we need to focus on care to ensure speedy recovery and maximum satisfaction. Hospitality is critical in healthcare as the patient and his/her attendants are already distraught and highly anxious. Hospitality is driven mainly by human interventions—in nursing, housekeeping, as well as food and beverages. It is very difficult to ensure consistency of quality and hence, we want to benchmark these to ensure we provide the best quality of hospitality all the time.
–Dr. Ananth Rao, Head–Quality Department, Apollo Hospital, Bangalore (December 2012)
Dr. Rao believed that although clinical services formed the core of Apollo’s services and brand image, hospitality would support the brand, and in the long run, both would merge to form the Apollo brand (as shown in Exhibit 3).
All services that did not require core clinical expertise were classified as hospitality services, including services such as billing, dietician service, food & beverages, facility, housekeeping, nursing, facility, and overall operations. Each service was executed through a variety of processes. All the processes included in each service were identified and defined with regard to the procedure, timelines, required output, and so on. All the processes were mapped and the quality measures defined; these would be used as Sigma metrics.
11
12
Express Healthcare, (2010), http://healthcare.financialexpress.com/201009/strategy01.shtml
Source: http://kolkata.apollohospitals.com/newsroom/271-apollo-hospitals-only-healthcare-super-brand-in-india.html
21
Page 4 of 20
Apollo Hospitals: Differentiation through Hospitality
The Quality Department, established under the leadership of Dr. Rao, comprised two dedicated staff—Soumi Dutta and Nisha Maria—who looked after a variety of quality-related issues. Soumi and Nisha collected feedback from the patients between March 2011 and December 2012using the form presented in Exhibit 4. Patients were asked to rate each department on a scale of 1 to 10. Additionally, open-ended feedback such as patients’ comments, opinions, or suggestions was also collected.
A schedule was developed for collecting feedback, which ensured that Soumi and Nisha collected feedback from a cross-section of patients; this also ensured that no biases crept into the feedback. The feedback collection methodology is shown in Exhibit 5. Soumi and Nisha were trained to collect frank, free-flowing feedback from the patients. If they received complaints while collecting feedback, they would immediately inform the department concerned and get the errors rectified, whenever possible; or ensure that the complaints were addressed to the patient’s satisfaction in real time. One of them recollected the following anecdote:
A patient had complained that the door was not getting locked properly. I got in touch with the facility personnel and they worked on the door and the lock and fixed the problem—all in a matter of 25 minutes from the moment it was brought to my attention. The patient was satisfied with the immediate solution. However, we did not stop there. We teamed up with the facility team and checked every door of the hospital and repaired them if required. We wanted to ensure that such complaints were not repeated.
The feedback collection process served multiple objectives such as collecting open-ended feedback from patients, resolving the issues in real time, and further auditing the actions of the service departments. The real-time escalation flowchart is shown in Exhibit 6.
The feedback was saved on an MS Excel spread sheet, and stored on a monthly basis for easy retrieval. The data was then analyzed using various parameters and trends were plotted for each service. At Apollo, each service was related to a department; hence, it was easier to deal with the complaints and determine monthly improvements.
FEEDBACK ANALYSIS
From March 2011 to December 2012, 1,434 complaints were received from among the 1,38,600 in-patients treated during that period (approximately 1.03%). A Pareto chart was plotted for these complaints (as shown in Exhibit 7).
The housekeeping department received the maximum number of complaints, while the dietary service had the least number of complaints. The department-wise spread of complaints is shown in Exhibit 7. Some of the complaints were genuine concerns while some were related to minor discomfort. A few of the complaints were very specific, while some were generic. All of these were analyzed, which would enable the hospital to work towards reducing the overall number of complaints. Some of the sample complaints from each department are provided in Exhibit 8.
According to Dr. Rao,
Every complaint is an opportunity to improve. We keep looking for the smallest of the complaints, which will help us in improving our quality by several levels. Sometimes it is difficult to interpret the complaints and it is even more complex to develop strategies that will enable a better patient experience. The complete data set was analyzed to determine the word frequency count in the complaints section. The significant words with their frequencies are shown in Exhibit 9. This analysis was used to focus on specific tasks to ensure reduction in the number of complaints. For example, the most significant word was “time” and it was associated with delays in response time for the various services. The twenty-fourth most frequent word was “late,” which is again related to response time. Thus, the word frequency technique helped in focusing on problem areas.
Based on the results of the analysis, benchmarks were set in consultation with the respective department for the response time of each service as shown in Exhibit 10.
Apart from this quantitative analysis, another approach was used to analyze the feedback and obtain deeper insights for quality improvements. Dr. Rao used the term “defect-defective” from the Six Sigma methodology—one
“defective” product/service could be caused by one or several “defects”. According to Dr. Rao,
22
Page 5 of 20
Apollo Hospitals: Differentiation through Hospitality
Any complaint from a patient is considered as a ‘‘defective’’. For example, consider the complaint: ‘‘Food is not served on time’’. This complaint may arise due to several reasons such as food not being prepared in time, food not being delivered on time, patient changing his/her order, etc. It is essential to identify these defects in order to eliminate the defective.
On receiving a complaint from the patient, which was termed as “defective,” defects that led to the defective
(complaint) were identified. Root-cause analysis was performed on all the processes of the identified defects. The processes were re-engineered to eliminate all the defects and a pilot study was conducted using the “Define Measure
Analyse Improve Control” (DMAIC) cycle. Once the process was found acceptable, it was then deployed across locations. This was followed by routine and surprise audits to ensure that the process was being followed as defined to ensure customer satisfaction. The flowchart is shown in Exhibit 11a and b. All feedback related to medical services was escalated to the Medical Director’s office.
In addition to this, the Quality Department at Apollo Bangalore developed a methodology called the Daily Point
Average© or DPA©. The ratings provided by patients for different departments were used to calculate the DPA©. The departments had to improve these ratings over a period of time. The DPA© effectively captured the “mind of the customers” since the feedback was collected during the patients’ stay and not at the time of discharge.
BENCHMARKING OF HOSPITALITY
Hospitality required high human involvement and was very specific not only to local cultures but also to each individual. Since a patient had to stay in a hospital to get cured, hospitality automatically came into the picture.
Hospitality in various hospitals was very different owing to the surroundings and differences in customer (patient) requirements. Patients did not walk into a hospital out of volition—they came in only because there was some problem. Under such conditions, the patient would be very agitated and any small thing that was out of place would become a big issue. Any kind of delay would be extremely intolerable and all the services had to be perfect all the time. Even in the hotel industry, hospitality was not standardized and benchmarks were not available. The Ritz–Carlton hotels, which are considered the gold standard in the hospitality industry, had used Six Sigma and benchmarking for their hospitality business.13 Although benchmarks for clinical services were well-established, those for hospitality in hospitals were yet to be established.
Apollo Bangalore developed benchmarks for several common complaints with three levels of services (as shown in
Exhibit 10) by adopting the Kano model, which was developed by Noriaki Kano (Exhibit 12). This model was used across service industries and it helped in understanding customer expectations from any product or service. The threshold or the basic quality was the minimum requirement of the customer, which would be taken for granted even if it were present; however, if it were not there, the customer would complain about its absence. Normal or performance quality was something that the customer would expect because these were stated either by the product/service provider or by the customer as a requirement. This quality was observed by the customer and its absence would cause discomfort and disappointment. Exciting quality of the service or product was something that would not disappoint the customer; the presence of this would delight the customer, since the customer did not expect this quality. With time, the exciting quality would become performance quality and the performance quality would become a basic quality. Hence, the manufacturer or the service provider should always strive to provide new exciting qualities.
Accordingly, several metrics were defined for benchmarking. For example, patients were informed that routine hospital-provided meals would be served within 10 to 20 minutes of every mealtime. This became a performance attribute. The threshold requirement of the patient would be that meals should be served within 20 minutes after placing the order. If the meal was served within 10 minutes, the patient would be delighted. However, if this customer (patient) were to come to the hospital again, she/he would expect the meal to be delivered within 10 minutes; this then would become a performance quality for her/him.
13
Source: http://www.qfdi.org/newsletters/six_sigma_qfd_hotel_application.html
23
Page 6 of 20
Apollo Hospitals: Differentiation through Hospitality
FINANCIAL IMPLICATIONS
The Quality Department required funding for data collection, analysis, and other related activities. These expenses affected the bottom line of the hospital. However, from March 2011 to December 2012, the number of re-visits increased, which implied that customer loyalty had increased. There was a 15% increase in the number of inpatients. Earlier, high discounts had been offered to dissatisfied patients owing to errors in service or poor quality of service. Gradually, there was a reduction in the discounts provided, which was a direct result of better satisfied patients. Additionally, owing to better processes, cost of quality (in terms of re-work and consumable wastage) had reduced, which helped in improving the bottom line. Further, owing to better service and higher levels of satisfaction, the patients acted as brand ambassadors for Apollo and provided word-of-mouth publicity, which improved the top line.
Another example of decreased turnaround time and a resultant increase in profitability was seen in the Biochemistry
Lab at the Apollo Bangalore Hospital. Dr. Rao headed this lab and he understood the patients’ requirement of receiving diagnostics reports in two hours instead of three. Dr. Rao and his team redesigned the process using 5S and lean concepts and managed to reach a turnaround time of two hours. Profits from the Biochemistry Lab nearly doubled after the decrease in turnaround time; while the cost of consumables increased by only 11%.
Even though the Apollo team was trying to improve customer satisfaction, it still faced the question of how much satisfaction could be actually provided to the customer considering the room tariffs that were charged. As seen from
Exhibit 13, the charges at the Apollo Bangalore Hospital ranged from USD 25 for a basic room to USD 120 for the
Platinum Suite. A Ritz–Carlton basic room would cost USD 799 at Washington, U.S.A. and USD 165 at Kuala
Lumpur, Malaysia (per person, per night). 14 The Apollo team might be able to provide high quality hospitality to patients in the Platinum Suites. However, the aspiration to provide the same service to patients in other rooms might not be financially feasible. The team was trying to build high levels of service for the Platinum Suites. However, the volumes in the other rooms were too high to be ignored, especially in the Indian context. Additionally, customer loyalty was extremely important to Apollo; in Dr. Panyala’s words,
Customer loyalty and not mere retention is what we need to focus on. It is important to think ahead of the customer to identify issues that may compromise the experience.
QUANTIFYING HOSPITALITY ACROSS APOLLO
According to Dr. Rao,
Once, we develop the benchmarks and the Sigma metrics, we want to replicate the system across all Apollo hospitals in the country. Each hospital will have to devise its own benchmark and
Sigma metrics. However, we want to provide a framework for developing these and then measuring the outcomes. All the hospitals would then be compared by equalisation of scores and would benefit from one another’s learning”.
After collecting the feedback and attempting to set benchmarks, Dr. Rao knew that he needed to go deeper and analyze each service through the complaints, set up relevant benchmarks, and target certain Sigma levels for each benchmark. He wondered whether they could collect and analyze data in a better manner. He wanted to arrive at the basis for the cost-benefit analysis of this activity. Looking at the complaints and the analysis, Dr. Rao had two major questions on his mind:
1.
2.
14
What strategy should be used to reduce the number of complaints and sustain the culture of excellence at Apollo Hospitals, Bangalore under the leadership of Dr. Panyala?
Given the manual intensive processes involved in addressing the hospitality issues, what is a good
Sigma level? Could Apollo set a target for Sigma level in hospitality?
Source: www.ritzcarlton.com, accessed on April 15, 2013.
24
Page 7 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 1
ACE@25 parameters
Sl.
Parameter
No.
1 Coronary artery bypass grafting (CABG) mortality rate
2 Complication rate post coronary intervention(percutaneous transluminal coronary angioplasty; PTCA)
3 Average length of stay (ALOS) post angioplasty
4 Average length of stay (ALOS) post total hip replacement (THR)
5 Average length of stay (ALOS) post total knee replacement (TKR)
6 Complication rate for total knee replacement (TKR)
7 Average length of stay (ALOS) post renal transplant
8 Average turnaround per dialysis chair per day
9 Average length of stay (ALOS) post transurethral resection of the prostate(TURP)
10 Complication rate transurethral resection of the prostate(TURP)
11 Endoscopy complication rate
12 Patient satisfaction with pain management
13 Door to thrombolysis time in ischemic stroke in emergency room (ER)
14 Percentage conversion of coronary angiographies to coronary artery bypass grafting (CABG)
15 Catheter-related blood stream infection (CR-BSI)
16 Ventilator associated pneumonia (VAP)
17 Catheter-related urinary tract infection (CR-UTI)
18 Average length of stay (ALOS) in hospital
19 Average length of stay (ALOS) in intensive care unit (ICU)
20 Door to CT time in stroke cases in emergency room (ER)
21 Surgical site infection (SSI – Clean wound)
22 Medication errors
23 Average length of stay (ALOS) post modified radical mastectomy (MRM)
24 Average length of stay (ALOS) post microdisectomy
25 Average urea reduction ratio*
26
Percentage of patients achieving/maintaining haemoglobin level of 11gram or higher after 3 months of dialysis in end stage renal disease (ESRD)
*Optional
25
Page 8 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 2
Accreditation of Apollo Hospitals
Accreditation
Joint Commission International (JCI)
Delhi,
Apollo Hospital Location
Chennai, Hyderabad, Ludhiana,
Bangalore,
Kolkata, Dhaka
National
Accreditation
Board
for
Hospitals
&
Madurai, Chennai
Healthcare Providers (NABH)
National Accreditation Board for Laboratories (NABL)
Chennai
ISO 9002
Chennai
Source: Apollo Investor Presentation (retrieved from www.apollohospitals.com in January 2013)
Exhibit 3
Apollo Brand–Clinical and Hospitality Services
Core Clinical
Services
Hospitality
Source: Interview with Dr. Ananth Rao
26
Page 9 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 4
Feedback Form
YOUR FEEDBACK Thank you for choosing Apollo Hospitals for your healthcare needs. As a quality improvement initiative, we are looking for improvements in parameters towards ‘‘Service Excellence’’ of our hospital.
Please provide a few minutes of your valuable time for a personal interaction.
How satisfied are you with your experience and the services provided by our hospital on a scale of 1 to 10?
1. MEDICAL SERVICES
1
2
3
4
5
6
Poor
2. NURSING SERVICES
1
2
3
7
8
Good
4
5
6
3. OPERATIONS &
ADMINISTRATION
1
2
3
8
5
6
7
1
2
3
1
2
4
3
5
6
7
1
2
3
4
5
Excellent
8
9
10
6
7
Excellent
8
9
Good
4
5
Poor
6
7
8
Good
Patient Name (Optional):
UHID:
Date of Admission:
Signature:
Date:
27
10
Excellent
COMMENTS (OVERALL):
Room No:
10
Good
Poor
6. FACILITY &
MAINTENANCE
9
Good
Poor
5. HOUSEKEEPING
SERVICES
10
Excellent
8
Poor
4. FOOD & BEVERAGES
9
Good
4
10
Excellent
7
Poor
9
9
10
Excellent
Page 10 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 5
Feedback collection methodology
A Typical Process Map: DPA
Daily Data Collection Method
Score given to IT
Dept. & disseminated to individual stakeholders via SMS by 3:00pm
1:00pm–2:30pm
Data
consolidated;
DPA Score developed* 11:00am–1:00pm
Survey conducted
Point of Data
Colle
ction
Details of complaints mailed to individual stakeholders by
3:00pm
th
15 day consolidated
DPA Score calculated
Data submitted to CEO & DMS
Data submitted to CEO & DMS
Patients staying for 1 day or more Discharged patients Monthly
Analysis
*The lower the score, the better
28
Quarterly
Analysis
Sustenance
Audits
Page 11 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 6
Flow for real-time complaint escalation
Survey conducted
Real-time escalation of issues within 15 minutes to stakeholders concerned
Resolution turnaround time (TAT) given by respective stakeholders Closure audits conducted at the end of TAT
Patient surveyed
NO
Issue rectified? YES
Issue closed
Plan preventive action
29
Stakeholder/representative meets patient for better understanding of issue Page 12 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 7
Department-wise complaints
HK
Department
No. of Complaints
373
F&B
318
Facility
236
Operations
189
Nursing
173
Medical
92
Billing
34
Dietary
19
30
Page 13 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 8
Sample complaints from each department in Apollo Hospitals
Housekeeping: Sample Complaints
Housekeeping service needed to be improved.
The dustbin was broken. The patient’s relative went to the housekeeping desk and got a new one, but housekeeping staff came and took the new dustbin away and replaced it with broken dustbin.
Linen was not changed in 24 hours. Today morning, when we requested for fresh linen, they said it would be provided in 5 minutes; it was not changed till 3:15 pm.
The TV remote was not provided in the room. We asked for it 3 to 4 times, but there was no response.
The toilet was stinking. Yesterday evening, the floor was dirty.
Drinking water was not provided when we asked for it, so finally we had to get it from home.
Housekeeping came to clean only once a day. The inpatient guide book said that they would come twice a day.
The clock battery had to be replaced.
Yesterday, I did not get a blanket the entire day.
It took 20 minutes to get a blanket.
Food & Beverages: Sample Complaints
The food served was not hot. I asked for kichidi but got only rice.
Around 8:30 pm last night, the patient asked for milk. We were told it would be served it by 9 pm, but the patient did not get milk. Today, we asked for milk again.
The food served was too spicy and dry. The food served was not what we had ordered.
The food served was not tasty, so we brought food from home.
The quantity of food served was not sufficient. The quality of roti was bad—it was dry and hard.
Coffee and milk were always served cold, even after complaints.
Food quality was not good
It took 1.5–2 hours for the plates taken away.
The cafeteria menu card was spoiled. They should get a new one.
Despite the patient informing the staff that she did not eat food with garlic or onions, the food served contained garlic and onions. After endoscopy was done, it took 1 hour for the food to be served despite repeated reminders.
There was too much pepper in the child’s food.
Facility: Sample Complaints
The nursing call bell did not work properly. The nursing call bell wire was short and did not reach the patient’s bed.
The room was too congested and stuffy. The temperature in the room should be checked regularly.
In the semi-private room, the light from the next room was too bright; the glass partition should be covered with curtains. The TV remote was taken away since the TV disturbed the other patient.
31
Page 14 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 8 (Contd.)
The staff members in the parking area were not cooperative. They were very arrogant. Last time, the patient was admitted to the ER at 11:30pm, but we were not allowed to park even for a few minutes.
The balcony was not clean. When we felt suffocated in the a.c. room, we opened the window. There was stagnant water outside, which led to a lot of mosquitoes coming in.
Hot water was not available in the bathroom.
Extra curtains should be provided. There were hooks for extra curtains. I had asked for curtains, but they were not provided. This hampered my privacy. The door made too much noise, which disturbed my sleep.
The flush was not working properly.
The attendant’s cot size should be a little bigger.
Operations: Sample Complaints
The doctor advised the patient undergo bronchoscopy; and the patient was asked not to eat anything from 8:30 am that day. At 12:30 pm, when the patient’s attender enquired with the bronchoscopy department, the attender was told that the procedure was postponed to the next day. No communication was provided to the patient regarding this; information was given only when the attender specifically asked for it.
The patient’s attender was told that the patient would be shifted to the ward from the OT recovery only after 1 hour, so the patient’s attender went for lunch. However, the patient was shifted in 20minutes.
The insurance staff left at 3:00pm on Sunday and there was nobody to assist us with the formalities.
We came to the Emergency Room at 10am. We were told we would get a room around 2pm. The room was finally allotted at 9pm.
A Telugu translator was required.
A “silence please” board should be put in each room.
The transport boy’s shoes were stinking. The hygiene of shoes should be improved.
All the staff should close the door while leaving the room.
We had to wait for the discharge summary as it was not updated on time.
The patient or the attenders should be made aware of what to bring to the hospital. When we asked the doctor what all we should bring, the doctor told us to get some clothes for the baby. Nobody informed us about the other things that would be required and were not included in the package. Information and orientation should be given beforehand and not at the time of an emergency.
Three days ago, while making an appointment over the phone, the receptionist made me wait for 3–4 minutes the first time. When I called again, the same person made me wait for 7–8 minutes. The appointment was finally given only when I called the third time.
32
Page 15 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 8 (Contd.)
Medical: Sample Complaints
The patient was in severe pain.
We expected the doctor to visit the patient in the morning. He had not come on rounds till 3 pm. We were waiting for the treatment plan and discharge.
Many staff members were seen walking into the canteen wearing OT gowns and shoe covers.
The operation theatre was available earlier, but was not allotted due to coordination issue (the surgery was scheduled at 12pm,but the patient was not shifted even at 3 pm).
The patient was suffering from pain in the stomach.
Backup treatment should be available for oncology patients in the ER and even in the ward.
Nursing: Sample Complaints
The nurses should attend to patients more frequently, especially on Sundays and other holidays.
HDU: there was a slight delay in service. Some of the nursing staff members were not prompt.
Blood sample had already been taken in the morning, but the staff came and took the sample again; they said that the previous blood sample could not be used as the blood had clotted. The sample collection process was extremely painful. Yesterday night, around 11:30pm, the nurse on duty pulled out the IV line in a rough manner, which caused a lot of pain, even though the patient kept asking the nurse to do it slowly.
The response time of nurses was bad. It took them at least half an hour to come after being called 2–3times.
It would be nice to have a specific set of nurses to take care of me regularly.
The nursing staff seemed to be overworked.
A checklist/patient education material for the plan of care should be available for patients’ convenience.
The staff should try to find out the patient's needs. The response time was very high. There should be additional staff, as the staff seemed to be overburdened.
Today morning, the nurse tried thrice but could not find a vein. She then called another nurse, who completed the procedure. My hand hurt for a long time after this.
Billing: Sample Complaints
Billing took a long time. It should be monitored and abnormal delays should be avoided.
During admission, the credit cell department staff behaved very rudely.
The discharge process took more than 2 hours.
What all were included in the package was not very clear to us.
The cost of surgery should be informed beforehand, rather than calling us at short notice and telling us to pay immediately. If we had been informed earlier, it would have been easier for us to arrange the finance and settle it.
We had to wait for 40 minutes for the finalization of the bill.
33
Page 16 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 8 (Contd.)
Dietary: Sample Complaints
Food should be less spicy and should be according to the child's taste.
The dietician did not visit the patient until 4:10pm. The patient was admitted on 9/6; the surgery (knee replacement) was performed on 10/6. The patient would be discharged the next day, and we needed to meet the dietician.
The doctor suggested normal diet but yesterday morning, the dietician changed it to soft diet. In the evening, the doctor changed it to normal diet. However, soft diet was continued until lunch today. After we complained, the F&B boy changed it to normal diet, but I had already finished my lunch by then.
The doctor told us not to give chutney and sambar to the patient. After having the food suggested by the dietician, the patient had stomach pain.
The dietician has not come to my room yet.
Exhibit 9
Text analytics–word frequency count
Sl.
No.
1
Word
Time
Count
271
Sl.
No.
19
Word
Good
Count
64
2
Food
220
20
Response
62
3
Nurse
214
21
Proper
58
4
Clean
206
22
Said
57
5
Toilet
139
23
Hot water
54
6
Water
129
24
Late
48
7
Doctor
107
25
Bedsheet
47
8
Improve
106
26
Quality
46
9
Wait
98
27
Floor
43
10
Staff
95
28
Serve
43
11
Bed
94
29
Lunch
41
12
Morning
94
30
Response time
41
13
Night
87
14
Attend
79
15
Work
79
16
Hot
78
17
HK
69
18
Minute
66
34
Page 17 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 10
SN KPI DEPT
1
F&B
Benchmarks for hospitality developed by Apollo Hospitals, Bangalore
MEASUREMENT
ROUTINE MEAL DELAY
COLOUR
TARGET
CODE
PROCESS
RED
>20MINS
YELLOW 10-20MINS
GREEN <10 MINS
FREQUENCY
FORTNIGHTLY
IN BETWEEN ORDERS DELAY
FOOD
RED
YELLOW
GREEN
>20MINS
5-20MINS
<5 MINS
WEEKLY
IN BETWEEN ORDERS DELAY
BEVERAGES
WRONG DIET
2 DIETARY
POST ADMISSION DIETICIAN
CONSULTATION DELAY
RED
>30MINS
YELLOW 15-30MINS
GREEN 5-15MINS
RED
NA
YELLOW
NA
GREEN
0
RED
DAILY
>120MINS
YELLOW 60-120 MINS
GREEN <60 MINS
3 FACILITY
WEEKLY
DAILY
REPAIR TIME
MINOR
RED
>30 MINS
YELLOW 15-30 MINS
GREEN <15 MINS
WEEKLY
REPAIR TIME
MAJOR
4
HK
5 NURSING
RESPONSE TIME
RESPONSE TIME
RED
YELLOW
GREEN
RED
YELLOW
GREEN
RED
YELLOW
GREEN
35
>60 MINS
30-60 MINS
<30 MINS
11-15 MINS
6-10 MINS
0-5 MINS
>3 MINS
1.5-3 MINS
0.7-1.5 MIN
DAILY
DAILY
DAILY
RED
YELL
KPI
GREEN
OW
SCORE
Page 18 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 11a
Defective/Defects analysis and process re-engineering flowchart
Patient complaint/feedb ack
Identify defect(s) leading to defective Defective
Root
Cause
Analysis
Repeated
Complaints
Process Changes
(Reengineering)
Pilot Study
(PDCA)
If solution acceptable Sustenance Audit
Deploy at all locations 36
If solution not acceptable
Page 19 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 11b
Defective/Defects analysis and process re-engineering flowchart
Delay in food service
Patient did not get food within 20 minutes Detailed
Root
Cause
Analysis
Delay due to various touch points involved
Solutions proposed
Repeated
Complaints
Direct phone line provided to F&B which can be contacted 24*7 from all patient rooms
Manager on duty to follow up Pilot Study
(PDCA) for private rooms
Sustenance
Audit
Deploy at all the wards 37
Solution
Accepted
Page 20 of 20
Apollo Hospitals: Differentiation through Hospitality
Exhibit 12
Exhibit 13
Kano Model
Room tariffs at Apollo Bangalore (as of January 2013)
Bed Type
Charges (USD)
Shared Bed Category-1
25
Shared Bed Category-2
36
Twin Sharing Bed
48
High Dependency Unit (HDU)
48
Private Bed
77
Executive
92
Platinum Suite
120
Intensive Care Unit (ICU)/Critical Care Unit (CCU)
80
38
4188
REV: NOVEMBER 30, 2011
V. KASTURI RANGAN
SUNRU YONG
Soren Chemical:
Why Is the New Swimming Pool Product Sinking?
Jen Moritz grimaced as she reviewed the February 2007 sales report for her company’s new
Coracle product. In September 2006, Soren Chemical had launched Coracle, a new water clarifier for use in small recreational and household swimming pools. Moritz was responsible for marketing the new clarifier, which she believed was a superior product, but the results so far were discouraging.
The volume target was 50,000 gallons (100,000 units) for the first year of sales. However, through the first half of the selling season for pool chemicals, Soren had sold just 3,725 gallons, or 7,450 units.
Moritz also had responsibility for marketing Kailan MW, a clarifier used primarily in large recreational water park facilities with typical capacities of one million gallons or higher. Very small quantities of Kailan MW were sufficient to treat large volumes of water, but it was unsuitable for smaller-scale applications such as residential pools. Thus Soren Chemical had developed Coracle, targeting smaller pools with a lower volume of swimmers (known in the industry as “bather loads”) and a less intense maintenance program.
In 2006, Kailan MW revenues were $6.1 million and sales were on pace for a 7% increase in 2007; it was healthy growth in a relatively mature market. Coracle had been budgeted at $1.5 million in sales for the year, but so far Soren had sold a very disappointing $111,000. The company’s internal analysis estimated a $30 million market in the United States for large-scale, commercial-use clarifiers such as Kailan MW. Moritz was convinced that an even larger market existed for Coracle. She explained her challenge:
How do we convince distributors to push Coracle and retailers to create shelf space for it?
Is the product priced so that the economics are attractive to our channel partners? From the end-user perspective, most pool owners don’t fully understand the safety and cost-saving benefits of Coracle. Should we invest aggressively to create greater awareness?
________________________________________________________________________________________________________________
HBS Professor V. Kasturi Rangan and writer Sunru Yong prepared this case solely as a basis for class discussion and not as an endorsement, a source of primary data, or an illustration of effective or ineffective management. The authors thank Simon Roy of Zodiac Marine & Pool Co. and
Brenda Barnicki of Eastman Chemical Co. for their valuable contributions to the development of this case.
This case, though based on real events, is fictionalized, and any resemblance to actual persons or entities is coincidental. There are occasional references to actual companies in the narration.
Copyright © 2010 Harvard Business School Publishing. To order copies or request permission to reproduce materials, call 1-800-545-7685, write
Harvard Business Publishing, Boston, MA 02163, or go to http://www.hbsp.harvard.edu. No part of this publication may be reproduced, storedin a retrieval system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the permission of Harvard Business Publishing.
Harvard Business Publishing is an affiliate of Harvard Business School.
39
4188 | Soren Chemical: Why Is the New Swimming Pool Product Sinking?
Company Background
Timothy Soren founded the company in 1942 to sell industrial-strength cleaning solutions. In the decades since, the company had expanded its focus to include industrial chemicals for lubricants and fuels, as well as a range of chemical solutions for treating drinking water and wastewater. In 2006, the Soren Chemical product line included over 350 products, and company revenues were $450 million (see Exhibit 1 for financial data).
Historically, Soren Chemical had concentrated on business-to-business sales and placed little emphasis on creating consumer awareness of its products. However, in 2002 the company had begun to invest selectively in developing brands for products that had potential in the consumer market.
While the company had enjoyed only modest success with its branding efforts, Soren Chemical continued to invest opportunistically in products with potential beyond the traditional “B2B” core.
Jen Moritz was a marketing manager in the Water Treatment Products group with responsibility for chemicals used in drinking and pool water treatment. Soren had identified in the pool water clarifier market a significant opportunity to build a consumer brand, and Moritz’s role was to develop the goto-market strategy.
Product Background
Kailan MW and Coracle are “flocculants,” chemicals that cause suspended particles in liquids to agglomerate into larger, heavier particles called “flocs.” They are particularly effective for microscopic particles that cannot be removed through physical filtration alone. Used under the appropriate pH, temperature, and salinity conditions, flocculating agents react with water to form insoluble hydroxides that physically trap small particles into the floc. The larger floc can then be trapped through sedimentation or filtration processes. Moritz explained:
Flocculants can be thought of as chemical “nets” that work in conjunction with filters. To get water clean you need to get rid of particles that would otherwise pass right through filters.
These particles can be as small as 0.5 microns. Only a flocculant can get something so small— the chemical “net” literally traps the little particles that filters miss, and then the net itself gets stopped by the filter.
In 2002, the Soren research and development team created the Kailan MW flocculant and found that it had application as a “clarifier” for pool water. This meant that Kailan MW was effective in dealing with oils that come from deodorant, hair spray, and lotions—which cause odors and cloudiness in water. Cloudy water is a safety hazard, obscuring pool depth for divers and making it difficult for lifeguards to see swimmers below the surface. What set Kailan MW apart from other pool clarifiers was its ability to combat organic debris. It could trap algae as well as dangerous waterborne pathogens such as E. Coli and cryptosporidium—a parasitic disease that normal chlorine levels do not kill. Moritz explained the importance of clarifiers as part of a pool maintenance regime:
The layperson knows only that pools have chlorine and filters. They do not realize there is a complex system of pumps, chemicals, and UV or ozone technology used to make the water safe and clean, with minimal odor. UV systems alone, for example, cannot effectively sanitize water when there is high turbidity or cloudiness. Kailan MW captures the smallest particles and quickly reduces the cloudiness. Similarly, chlorine alone is not sufficient to make water safe because there are resistant, waterborne illnesses such as cryptosporidium. Any large, public pool or water park can be prone to an outbreak. That is where Kailan MW comes in.
2
BRIEFCASES | HARVARD BUSINESS PUBLISHING
40
Soren Chemical: Why Is the New Swimming Pool Product Sinking? | 4188
One gallon of Kailan MW could effectively treat approximately 500,000 gallons of water. Most competing clarifiers required daily application, but Kailan MW’s advantage was its effectiveness over a longer period. Depending on the intensity of pool usage, operators could use it every other day, rather than the daily use required with similar products.
Commercial-Use Clarifiers Market
Soren Chemical estimated the 2007 U.S. market for specialty commercial-use clarifiers to be approximately $30 million, for which primary demand came from commercial pools (there are
300,000 in the U.S.) and water parks (1,000 in the U.S). Most commercial facilities must accommodate high bather loads and a disproportionate number of children, and it is imperative to maintain high levels of clarity and to protect against waterborne pathogens.
Soren Chemical sold Kailan MW for these commercial applications primarily through seven chemical formulators (see Figure A for an overview of the channel structure). The formulators produced or sourced the full range of pool chemicals, including chlorine tablets or liquids, cleansers, enzymes, shock treatments, and algaecides. Most products, including Kailan MW, were sold by the formulators under a private brand name.
Figure A
Pool Chemicals Channel Structure
Chemical Manufacturers
Formulators
Wholesale Distributors
Pool Specialty
Retailers
Mass Retailers
Pool Service
Professionals
Water Parks and
Commercial Pools
Residential Pool Owners
The formulators did more than simply deliver product; they typically provided a range of valueadded services, such as custom blending and packaging services tailored to customer specifications.
In some cases, they developed customized maintenance programs by working with water safety consultants, pump and filter manufacturers, as well as the chemical manufacturers themselves.
These programs included prescribed chemical regimes tailored for different commercial applications based on typical bather loads, filter types, and other cleaning technology. The programs also
HARVARD BUSINESS PUBLISHING | BRIEFCASES
41
3
4188 | Soren Chemical: Why Is the New Swimming Pool Product Sinking?
accounted for regional climate conditions that affected water quality, such as pollen, insects, and pollutants. Moritz explained Soren’s go-to-market strategy with Kailan MW:
These smaller formulators are able to provide custom blends and programs, which often make them preferred suppliers for commercial usage. Soren Chemical’s model does not allow us to customize for individual water parks and commercial pools. Since Kailan MW is bestsuited for this market, it made sense for us to partner with the formulators.
Residential Pool Clarifiers Market
Because Kailan MW was designed for large-scale commercial facilities, Soren Chemical did not intend it for use in smaller pools for fear of misuse and potential safety risks. However, in 2005 the sales team learned that at least two of its formulators had begun marketing a diluted version of
Kailan MW as a private label clarifier for the residential pool market. Moritz recognized that there was a significant, untapped market if the product could be appropriately refined for use in smaller residential pools. After several months of product development, the company unveiled a clarifier featuring chemical structure and properties similar to Kailan MW’s, but designed for smaller pools, lower bather loads, and less-frequent usage. Soren named the new product “Coracle,” after the traditional Welsh fishing boat that moved with the current, sweeping fish into its net.
The potential consumer market for Coracle was more fragmented than the commercial market for
Kailan MW. Industry reports showed that there were nearly 9 million residential swimming pools in the United States, all of which required regular maintenance and cleaning by their owners or a professional pool service. Most pool equipment and chemical manufacturers sold via a two-step process to wholesale distributors, which then supplied specialty retailers and service professionals.
Distributors typically carried tens of thousands of products from many suppliers—including the regional formulators that sold Kailan MW—and served a broad base of local retailers and pool maintenance companies. While there were a few national wholesale distributors, such as Pool
Corporation, many were small regional players. Some manufacturers bypassed the wholesale distributors altogether, selling a limited range of pool products directly to Wal-Mart or “do-ityourself” retailers such as Home Depot and Lowe’s. However, this approach was costly and was used only by the largest chemical companies.
Purchasing behavior differed between professionals and consumers. The professional market, comprising pool builders, cleaning and service companies, and independent contractors, typically bought supplies from the wholesale distributors. Industry experts believed there were 40,000 to
50,000 pool contractors and service companies in the United States. Approximately 80% of consumers maintained their own pools, and they generally purchased supplies from local specialty retailers or national retailers. Whereas the critical concerns with cloudy water in the commercial market were waterborne disease and swimmer safety, the consumer market emphasized aesthetics and perceived cleanliness.
Moritz identified three leading competitors for residential pool-use clarifiers: Keystone Chemicals,
Kymera, and Jackson Laboratories. She estimated that each of the companies had a 15% to 20% share of the residential pool clarifier market. Based on discussions with multiple distributors, she pieced together a product and pricing comparison (see Table A).
4
BRIEFCASES | HARVARD BUSINESS PUBLISHING
42
Soren Chemical: Why Is the New Swimming Pool Product Sinking? | 4188
Table A
Product Comparison
Product
Soren
Chemical
Coracle
Cost per container, retail price
Form
Gallons per container
Cost per gallon, retail price
Cost per ounce/tablet
Ounces/tablets needed per 30K gallons
Cost per treatment, retail price
Replacement cycle (no. of monthly treatments)
Monthly cost, retail price
Average months of pool usage
Annual cost, retail price
Keystone
Chemicals
Purity
Kymera
HydroPill
Jackson
Labs
ClearBlu
$25.00
Liquid
0.50
$50.00
$0.39
10.0
$15.00
Liquid
0.25
$60.00
$0.47
5.0
$3.50
Tablet
NA
NA
$3.50
1.0
$15.00
Liquid
1.00
$15.00
$0.12
32.0
$3.91
2.0
$7.81
5.0
$2.34
4.0
$9.38
5.0
$3.50
2.0
$7.00
5.0
$3.75
4.0
$15.00
5.0
$39.06
$46.88
$35.00
$75.00
Note: Costs are rounded.
Most service professionals considered ClearBlu to be the most effective chemical flocculant.
Neither Keystone’s Purity nor Kymera’s HydroPill had a discernable effect in reducing the need for chlorine, shock treatments, or enzymes. In contrast, industry experts believed that ClearBlu reduced the need for other chemicals by approximately 15%, although Jackson Laboratories did not emphasize this benefit. The key disadvantage of ClearBlu was the relatively high dosage required per treatment: a quarter-gallon of product was required each week. This was higher than the dosage required with the more concentrated products of competitors, making it less convenient for pool owners and service professionals to store ClearBlu in bulk.
Marketing Strategy for Coracle
A study by Soren Chemical showed that Coracle, like Jackson Laboratories’ ClearBlu, significantly boosted the efficiency of other pool chemicals by reducing the “burden” placed on chlorine. Moritz described the benefits of Coracle:
In a typical pool, two-thirds or more of chlorine is used up oxidizing organic contaminants such as oils. This keeps the chlorine from sanitizing the water, which really defeats its primary purpose. Coracle takes care of these oils, freeing up the chlorine to serve its primary function of fighting waste. It also reduces the need for shock treatments and enzymes.1
The potential cost savings for pool owners could be significant. The company’s research and development team estimated that Coracle reduced the need for additional chlorine, shock treatments,
1 Shock treatments or shock oxidizers are heavy doses of chemicals (typically chlorine) used in conventional pools to clean out
contaminants. Shock treatments may be required once a week or more, depending on the intensity of pool usage. Due to the high dosage of chemicals, shock treatments can make pools unusable for an hour or more. Enzymes help reduce organics and phosphates in water and are primarily intended to reduce “scum-line” formation at the edge of the pool, making it easier to clean. HARVARD BUSINESS PUBLISHING | BRIEFCASES
43
5
4188 | Soren Chemical: Why Is the New Swimming Pool Product Sinking?
and enzymes, reducing pool owners’ annual chemical costs by 20% to 30%.2 Soren Chemical made this benefit the central thrust of its marketing message, prominently featuring the claim on the bottle.
In order to reach service professionals and specialty retailers, Soren sold Coracle through wholesale distributors. The company allowed formulators to use private branding for Kailan MW and, indeed, most opted to sell clarifiers that did not explicitly identify Kailan MW as the flocculant.
However, the company decided not to allow private-label branding for Coracle, despite requests from several major wholesale distributors that marketed a number of pool chemicals under their own labels. Moritz explained how this fit into a longer-term plan:
In keeping with Soren’s new strategy to opportunistically develop consumer brands, our plan with Coracle is to make it a branded product. Our R&D pipeline includes other products, such as algae and phosphate removers that have potential in residential pools and spas. If we can build some recognition with the Coracle name, we have a platform for other products.
Ideally, Coracle will become almost a consumer packaged good that pool service professionals, specialty retailers, and consumers will know and request.
The manufacturer price for a two-quart bottle of Coracle was $14.88. For most products such as chlorine, shock treatments, and enzymes, distributors typically had a 20% gross margin. However, wholesale distributors expected to maintain a 30% gross margin in selling Coracle, as it was a differentiated chemical agent. Retailers and service professionals usually took a 15% gross margin, resulting in a suggested retail price of $25 to consumers (see Exhibit 2 for margin structure data).
The product launch in fall of 2006 was accompanied by a new website and a press release in three trade association journals targeted at pool service professionals and specialty retailers. The announcement emphasized Coracle’s performance advantage in trapping dangerous waterborne pathogens such as E. Coli and cryptosporidium. Soren Chemical received over 2,000 inquiries from service professionals and retailers in the first three months of 2007. The company responded directly with a brochure, technical notes, and material safety data sheet.
Despite these efforts, sales through February were a dismal $111,000.
The Dilemma
Frustrated with such underperformance, Moritz reviewed her marketing plan with colleagues, noting that Soren Chemical’s relative inexperience with marketing consumer-oriented brands appeared to be hindering the efforts with Coracle. One colleague questioned whether Moritz had fully accounted for competing products:
We are new to the residential pool market and perhaps we do not fully understand the consumer. How is Coracle beneficial to them and do they understand that? They already have other choices, and Coracle is around $25, which is higher than the unit prices of competitors. I think we need to use the pricing of comparable products as our guideline.
A second colleague pointed out the challenges of selling through distributors and retailers:
2 Cost savings estimates are based on chemical usage for conventional pools, which comprise over 85% of total swimming pools in the U.S. Salt water pools use significantly less chemicals than conventional pools, although clarifiers provide similar water safety benefits.
6
BRIEFCASES | HARVARD BUSINESS PUBLISHING
44
Soren Chemical: Why Is the New Swimming Pool Product Sinking? | 4188
We need to be certain the economics are attractive for distributors and retailers. After all, they already carry other clarifiers. Why would they make shelf space to carry Coracle? I worry that their incentives are not totally aligned with ours. If we have room to raise prices, it might create enough margin to make this more attractive to them.
To diagnose the situation more carefully, Moritz decided to conduct a survey among the pool service professionals and specialty retailers who had made inquiries about Coracle. The survey revealed that only 30% of the respondents recalled receiving the Coracle materials that Soren
Chemical had sent in response to their inquiries. Furthermore, though Soren Chemical had passed the contact information for interested customers to the appropriate wholesale distributors, nearly
70% of respondents stated that Coracle had not been offered by their distributors. Creating greater demand from service professionals could provide a jumpstart for Coracle, but Moritz estimated that she would need a $600,000 budget to conduct a mailing campaign and to run advertising in industry publications. Moritz also used the survey to learn from retailers and professionals about consumers’ use and understanding of clarifiers (see Exhibit 3). Soren Chemical had priced Coracle aggressively at a $25 suggested retail price, and the estimated $39 annual cost for the average pool was lower than with most competing products. Moritz suspected that most residential pool owners simply did not realize the value of Coracle relative to other clarifiers.
The selling season for residential pool chemicals would be largely over by May. Moritz needed to reevaluate the entire go-to-market strategy. If she could identify the problem, perhaps she could partly salvage 2007 and position Coracle for a successful second year.
HARVARD BUSINESS PUBLISHING | BRIEFCASES
45
7
4188 | Soren Chemical: Why Is the New Swimming Pool Product Sinking?
Exhibit 1
Soren Financial Data ($ millions)
2004
2005
2006
$ 439
335
104
$ 444
339
105
$ 450
345
105
Selling, General, & Administrative Expense
Research & Development
Depreciation
29
8
15
30
8
16
31
9
15
Operating Expense
52
54
55
Operating Income
52
51
50
Provision for Tax
Net Income
12
40
11
40
12
38
121
36
$ 15
130
39
$ 17
134
40
$ 17
Cost per
Container
Cost per
Treatment
Revenue
Cost of Goods Sold
Gross Profit
Water Treatment Group
Revenue
Gross Profit
Operating Income
Exhibit 2
Margin Structure
Retailer / Service Professional price
Retailer / Service Professional gross margin %
Retailer / Service Professional gross profit
$25.00
15%
$3.75
$3.91
15%
$0.59
Distributor price
Distributor gross margin %
Distributor gross profit
$21.25
30%
$6.38
$3.32
30%
$1.00
Soren Chemical price
Soren Chemical gross margin %
Soren Chemical gross profit
$14.88
35%
$5.21
$2.32
35%
$0.81
Exhibit 3
Specialty Retailer Survey, Selected Data
Annual chemical costs at retail prices, excluding clarifiers (recommended regime)
% of consumers who understand and use clarifiers regularly
Annual average cost of clarifiers at retail prices (recommended regime)
8
$300 average
25%
$50
BRIEFCASES | HARVARD BUSINESS PUBLISHING
46
Harvard Business School
9-599-114
May 25, 1999
Note on Behavioral Pricing
It is important for a firm to get its pricing “right.” Consider the impact of product pricing on a firm’s net income. If Coca-Cola could increase its prices by an average of 1%, without affecting consumer demand for its products, it would increase its net income by 6.4%. A price increase of less than 1¢ on a can of cola would translate to an increase in net income of about $300 million.
Similar 1% price increases, if they did not negatively impact demand, would lead to increases in net income of 16.7% for Fuji Photo, 17.5% for Nestle, and 26% for the Ford Motor Company. In fact, an average price increase of 1% would boost the net income of the typical large U. S. corporation by about 12%.1
These examples highlight the potential impact on a firm’s net income of optimally setting product prices — small changes in price can have an enormous impact on income. Before raising (or lowering) prices in an attempt to improve the bottom line, however, a firm must understand and anticipate a consumer’s response to a product price change.
Unfortunately, it appears that many firms lack the necessary understanding of a consumer’s
“willingness to pay” to optimally set product prices. When asked whether they were “wellinformed” on six of the potential inputs to the product pricing decision, managers at one wellrespected U.S.-based multinational responded as follows:2
•
•
•
•
•
•
84%
81%
75%
61%
34%
21%
were well-informed on the variable cost of providing their product. were well-informed on the fixed cost of providing their product. were well-informed on the price of competitors’ products. were well-informed on the value of their product to the customer. were well-informed on how consumers would respond to price changes. were well-informed on consumers’ willingness to pay at various price levels.
These managers were well-informed on the costs of providing its products and on the price of competitor’s products. They were also well-informed on the value its products delivered to consumers. However, when it came to a consumer’s willingness to pay or to a consumer’s response
1
2
From Robert J. Dolan and Hermann Simon’s, Power Pricing, The Free Press, New York, NY (1995)
Ibid.
Assistant Professor John T. Gourville prepared this note as the basis for class discussion rather than to illustrate either effective or ineffective handling of an administrative situation.
Copyright © 1999 by the President and Fellows of Harvard College. To order copies or request permission to reproduce materials, call 1-800-545-7685, write Harvard Business School Publishing, Boston, MA 02163, or go to http://www.hbsp.harvard.edu. No part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the permission of Harvard Business School.
1
47
599-114
Note on Behavioral Pricing
to potential price changes, these managers were lacking the insight needed to optimally set prices.
Experience suggests that this company is not unique in this regard.
Purpose of this Note
Any firm’s ability to optimally set product prices is governed by many factors. Some of these factors are well understood and are routinely incorporated into a firm’s pricing decisions. These factors tend to be heavily weighted toward those economic factors that are easily obtained by a firm, including their own variable and fixed costs of production, the market price of competitors’ products and the firm’s internal assessment of the value that their product delivers to the intended consumer.
Not surprisingly, these are the four factors on which the surveyed managers claimed that they were well-informed. We will briefly review these factors in a moment. 3
However, optimal product pricing also hinges on a consumer’s willingness to pay and on a consumer’s response to price changes, factors on which the surveyed managers claimed they were poorly informed. In addition, research has shown that a consumer’s willingness to pay is often influenced by “psychological” or “behavioral” variables that typically are not considered when setting price. Specifically, consumers often are as concerned with the behavioral question of “how fair a deal am I getting” as they are with the economic question of “how good a deal am I getting.”
This note is an attempt to highlight the potential impact of some of these behavioral variables on a consumer’s willingness to pay. In the process, it provides a more complete picture of consumer response to pricing and provides some insight for optimal product pricing.
Value Pricing and the Economic Perspective
The traditional economic approach to product pricing is driven by a small handful of factors, as shown in Figure A. One of these factors is the “objective value” the product delivers to the consumer. 4 This is a measure of the benefits that the product delivers to the consumer, regardless of whether the consumer recognizes those benefits. When 61% of managers claim they are wellinformed on the value of their product to the consumer, they are most likely referring to this
“objective value.”
A second factor in the economic approach to pricing is the “perceived value” of the product to a consumer. Perceived value is the value the consumer understands the product to deliver.
Sometimes, a product’s benefits are readily apparent to the consumer and “perceived value” approaches “objective value” with little effort by the firm. Other times, a product’s benefits are less obvious and need to be communicated by the firm to the consumer (e.g., via advertising, personnel selling). In such cases, the “perceived value” of a product typically falls below its “objective value.”
The perceived value of a product also can be influenced by the price of competing products or “substitutes.” Company A may develop a product that creates great objective value for consumers.
Consumers may recognize this value and be willing to pay a high price to obtain the product.
3
For a more complete discussion of these economic factors, you should refer to Professor Corey’s “Note on
Pricing” [HBS Note #580-091] or Professor Dolan’s “Pricing Policy” note [HBS Note #585-044].
4
Given consumer heterogeneity, “objective value” and “perceived value” will tend to vary across consumers.
For some consumers, these values will be high, for others, they will be low or zero. For simplicity, we will ignore consumer heterogeneity in this note and only consider the “typical” consumer.
Nevertheless, the behavioral perspective offered in this note apply equally well to a heterogeneous consumer population.
2
48
Note on Behavioral Pricing
599-114
However, if Company B introduces an identical product at a much lower price, the perceived value of Company A’s product would be reduced to the price of Company B’s product.
It is important to note that the “perceived value” of a product to a consumer should equal the maximum price that consumer is willing to pay for the product. Imagine a consumer who perceives the value of a modem to be $100. If priced above $100, the consumer has no incentive to buy the modem. If priced at $100 or less, however, the consumer always stands to gain from purchasing.
Figure A
Value Pricing and the Economic Perspective
Marketing
Efforts
Objective Value
Perceived Value
Consumer’s Incentive to Purchase
= [Perceived Value - Price]
Price of
Substitutes
Product Price
Firm’s Incentive to Sell
= [Price- COGS]
Cost of Goods Sold
$0
The last major component to the economic approach to pricing involves the firm’s cost of goods sold (i.e., COGS). Just as the consumer requires an incentive to purchase a product, the firm requires an incentive to sell the product. In order to stay in business and make a positive return, a firm must charge a price that covers both its cost of production. 5
All of these economic factors come together to form the “value pricing” approach to pricing.
In optimally pricing a product, a firm is bound at the upper end by the consumers’ “perceived value” for the product. This “perceived value” is influenced by the “objective value” of the product to the consumer, by the firm’s marketing effort to communicate that objective value, and by the price of substitute products. At the same time, the firm is bound on the lower end by its COGS.
By pricing above COGS and below perceived value, the firm has an incentive to sell the product, measured as [price – COGS], and the consumer has an incentive to purchase the product, measured as [perceived value – price]. In value pricing terminology, the firm has “created” value by offering a product that the consumer values at a price greater than the firm’s COGS. In turn, by pricing between perceived value and COGS, the firm has “captured” some of that value for itself and has allowed consumers to capture the remainder.
5
For simplicity, we will ignore strategic reasons for pricing below cost such as to build share or volume or to temporarily respond to a competitor’s pricing efforts.
3
49
599-114
Note on Behavioral Pricing
Adding a Behavioral Component to the Economic Perspective
This “value pricing” framework provides a basic model of how an economically rational consumer should respond to a firm’s pricing of a product. A rational consumer should purchase a product as long as the “perceived value” of that product is greater than the actual price being charged. In addition, the more one’s perceived value exceeds actual price, the greater should be a consumer’s incentive to buy. This leads to the fairly straight forward claim that:
The Economic Perspective: Consumers Buy When Perceived Value Exceeds Price
Consumers should purchase an item whenever the perceived value of that item exceeds its actual price [i.e., whenever (Perceived Value – Actual Price) > 0].
Adding a Behavioral Component
To complement this economic perspective, we now add a “behavioral” or “psychological” perspective to product pricing. This perspective captures “how fair a deal” one is getting. To make this point clear, consider the following scenarios first proposed by Professor Richard Thaler. 6
Scenario #1:
You are lying on the beach on a hot day. All you have to drink is ice water. For the past hour, you have been thinking about how much you would enjoy a nice cold bottle of your favorite beer. A friend gets up to make a phone call and offers to bring back a bottle of your favorite beer from the only nearby place where beer is sold — a small, run down-grocery store.
He says that the beer might be expensive and asks how much you are willing to spend. He says he will not buy the beer if it costs more than the price you state. What price do you tell your friend?
What is your “perceived value” for a nice cold bottle of your favorite beer brought to you on a hot beach? Once you have decided upon the price you would tell your friend, consider a second scenario, identical to the first, except for the source of the beer, which is underlined.
Scenario #2:
You are lying on the beach on a hot day. All you have to drink is ice water. For the past hour, you have been thinking about how much you would enjoy a nice cold bottle of your favorite beer. A friend gets up to make a phone call and offers to bring back a bottle of your favorite beer from the only nearby place where beer is sold — a fancy resort hotel. He says that the beer might be expensive and asks how much you are willing to spend. He says he will not buy the beer if it costs more than the price you state. What price do you tell your friend?
If you are like most people, your responses to these two scenarios differ. When Thaler presented these scenarios to a group of executives in the early 1980s, the median response in the grocery store scenario was $1.50 and the median response in the fancy resort hotel scenario was $2.65.
Similar results repeatedly have been obtained using Harvard MBAs — albeit, with somewhat higher average prices.
6
Scenarios 1 and 2 have been adapted from Richard H. Thaler’s paper, “Mental Accounting and Consumer
Choice,” Marketing Science, 4, 3 (Summer 1985): p. 199-214.
4
50
Note on Behavioral Pricing
599-114
The interesting question is why? As pointed out by Thaler, for the person consuming the beer on the beach, nothing of importance has changed between the two scenarios. Specifically,
•
in both scenarios, the ultimate consumption is identical — the same beer is consumed on the same beach.
•
no atmosphere from the fancy resort hotel or the run-down grocery store is being consumed by the beer drinker to justify different prices.
•
there is no strategic reason to report a price below one’s “perceived value” for the beer [e.g., you cannot haggle over price with hotel or store owner].
As a result, a person’s “perceived value” for the beer should be identical across the two scenarios. To report otherwise would suggest that a bottle of beer consumed on a beach somehow tastes better or quenches thirst more effectively when purchased from one location than another. In turn, if the “perceived value” of the beer should be identical across the two scenarios, a person’s
“willingness to pay” also should be identical across the two scenarios.
Yet people’s prices do differ and it appears that this difference is due to expectations consumers have regarding the price of a bottle of beer at a fancy hotel versus a run-down grocery store. As noted by Thaler, “While paying $2.50 for a beer is an expected annoyance at the resort hotel, it would be considered an outrageous ‘rip-off’ in a grocery store.”
In the end, it appears that one’s “willingness to pay” in these two scenarios is driven not only by the “economic utility” of the transaction [i.e., perceived value – price], but also by the
“psychological utility” of the transaction, driven largely by a consumer’s perception of “fairness.”
Over the next several pages, we will look at scenarios that highlight specific drivers of transaction
“fairness” and of the “psychological utility” of a transaction.
Some Behavioral Updates to the Economic Perspective
1.
The Relative versus Absolute Value of Money
In the economic approach to pricing, all money is equal – e.g., $10 in one transaction is worth the same as $10 in another transaction. Research suggests that this may not be the case when it comes to one’s willingness to pay, however. Consider the following scenario and think about how you would respond. 7 There are no right or wrong answers. There is only your intuition as to how you would behave if you found yourself in such a scenario.
Scenario #3:
You set off to buy a Sony Walkman at what you believe to be the cheapest store in the area.
Upon arriving, you find that the Walkman you want costs $29, a price consistent with your prior expectations. As you are about to make the purchase, a reliable friend tells you that the very same Walkman is selling for $10 less at a store approximately 10 minutes away. Do you go to the other store to buy the Walkman?
7
Scenarios 3 and 4 have been adapted from Richard H. Thaler’s paper, “Toward a Positive Theory of
Consumer Choice,” Journal of Economic Behavior and Organization, 1 (1980), p. 39-60.
5
51
599-114
Note on Behavioral Pricing
What would you do? Do you go to the other store and save $10 or do you simply go ahead and purchase the Walkman in the first store and pay $29?
A purely economic approach to this question would be to ask yourself whether 10 minutes of your time is worth $10. If you decide that 10 minutes of your time is more valuable than $10, you should forget about the potential savings and purchase the Walkman at the first store for $29. If, however, you decide that 10 minutes of your time is less valuable than $10, you should travel to the second store and purchase the desired Walkman there for $19.
Now consider a second scenario, identical to the first, except for the nature and the price of the product being purchased.
Scenario #4:
You set off to buy a Sony Camcorder at what you believe to be the cheapest store in the area.
Upon arriving, you find that the Camcorder you want costs $495, a price consistent with your prior expectations. As you are about to make the purchase, a reliable friend tells you that the very same Walkman is selling for $10 less at a store approximately 10 minutes away.
Do you go to the other store to buy the Camcorder?
From the economic perspective, if you decided to travel to the second store to save $10 in the first scenario, you also should have decided to travel to the second store to save $10 in this second scenario. In both scenarios, the tradeoff is $10 for 10 minutes of your time.
If you are like most people, however, your natural inclination will be to answer “yes” in
Scenario #3 and “no” in Scenario #4. After all, $10 on a $29 Walkman represents a savings of over
33%, but $10 on a $495 Camcorder represents a savings of a measly 2%. While the dollar savings are the same, the psychological value of the savings is far greater in the first scenario than the second.
Whereas a $10 savings on a $29 Walkman is perceived as a “fair” (and even generous) incentive to travel to the second store, a $10 savings on a $495 camcorder is perceived as a rather inadequate incentive to travel to the second store.
These two scenarios raise a curious and important fact about money. Namely, the
“psychological utility” of a fixed amount of money (e.g., $10) is relative. Saving $10 on a $29 item will have much greater impact on a consumer’s behavior than saving $10 on a $495 item.
This directly carries over to a consumer’s willingness to pay. Imagine two consumers, one of whom is debating whether to pay $19 for a Walkman that he values at $29, the other of whom is debating whether to pay $485 for a camcorder that he values at $495. While both consumers have the same “economic utility” to enter their respective transactions [i.e., perceived value – price = $10 in both cases], the first consumer will be more likely to make a purchase than the second due to the higher relative incentive to enter his transaction.
This leads to our first update to the economic perspective:
Behavioral Update #1: Willingness to Pay is Impacted by Relative Incentives
In determining his willingness to pay, a consumer will consider both his absolute
“economic utility” from the transaction [i.e., perceived value – actual price] and his relative incentive to enter the transaction [i.e., (perceived value – actual price)/(actual price)].
6
52
Note on Behavioral Pricing
599-114
2. The Impact of a Salient Reference Price
Expectations about “what a product will cost” also seem to impact a consumer’s “willingness to pay.” 8 Read the following scenario and think about how you would respond. Again, there are no right or wrong answers. Only your intuition matters.
Scenario #5
Your favorite sports team has made the playoffs. Its first-round playoff series is a best-ofseven series9 with Games 1, 2, 5, and 7 played on your team’s home field. General admission tickets had been priced at $20 during the regular season. The team decided to raise general admission prices to $40 for these four playoff games. Is this price increase fair or unfair?
Is it fair for a sports team to raise prices by $20 between the regular season and the playoffs?
Most people who encounter this scenario say “yes.” About two-thirds of Harvard MBAs not only believe it to be “fair,” but would expect such an increase. Arguments in support of this stance include: •
The nature of the product has changed – playoff games are more exciting than regular season games.
•
Demand almost certainly will be higher for the playoff games and the team is merely responding to this increase in demand.
•
It is common for ticket prices to increase for playoff games.
Now consider a second scenario, similar to the first scenario except for the timing of the $20 price increase.
Scenario #6
Your favorite sports team has made the playoffs. Its first-round playoff series is a best-ofseven series with Games 1, 2, 5, and 7 played on your team’s home field. General admission tickets had been priced at $20 during the regular season. General admission tickets were also priced at $20 for Games 1 and 2 of the playoffs. After Game 2, the team decided to raise prices to $40 for Games 5 and 7. Is this price increase fair or unfair?
In considering this second scenario, note that from a purely economic perspective, if Games 5 and 7 are worth $40 in Scenario #5, they should also be worth $40 in Scenario #6. Therefore, any difference in “fairness” between the two scenarios is not being driven by a change in the “economic utility” of the transaction, as measured by [perceived value – price], but by the “psychological utility” of the transaction.
8
In these and subsequent scenarios, individuals are being asked to assess the “fairness” of a firm’s pricing decision. Perceptions of fairness should impact “willingness to pay” in a straightforward fashion. In particular, in the short-term, consumers will be less willing to pay a price they feel is “unfair.” In the long-term, consumers will be less likely to purchase from a firm that makes an unfair pricing decision.
9
In a best-of-seven series, the teams play games until one team wins four. This guarantees a minimum of four games and a maximum of seven. Often, Games 1, 2, 5, and 7 are played at one teams location, Games 3, 4 and 6 are played at the other teams location. Also, for each of Games 5, 6 and 7, tickets often are not sold until it is apparent that each game will be required.
7
53
599-114
Note on Behavioral Pricing
If you are like most individuals, however, you find the price increase in Scenario #6 to be
“unfair.” What seems to drive this sense of unfairness is the fact that ticket prices were raised midplayoffs. Once ticket prices were set for Games 1 and 2, those prices created an expectation for prices for the remainder of the playoffs. By raising prices for Games 5 and 7, this expectation was violated.
These two scenarios highlight the power of a salient “reference price,” a price against which consumers compare other prices to assess both the “goodness” and “fairness” of a given transaction.
In Scenario #5, regular season ticket prices were not a salient (nor appropriate) reference price due to the fundamental difference between regular season games and playoff games. In contrast, in
Scenario #6, ticket prices for playoff Games 1 and 2 did establish a salient (and seemingly appropriate) reference price for Games 5 and 7. As a result, price changes in Scenario #5 are deemed
“fair” while price changes in Scenario #6 are deemed “unfair.”
Interestingly, Scenario #6 was actually played out in the spring of 1997. In a best-of-seven playoff series between basketball’s Miami Heat and New York Knicks, Miami raised ticket prices in the middle of the playoff series. For Games 1 and 2, Miami Heat management had set the prices for various seats at $20, $30 and $40. After Game 2, the Heat raised ticket prices to $50, $80 and $90 for
Game 5. Public outrage resulted and Game 5 was one of the very few basketball playoff games that year not to sell out. The extent of the outrage forced Miami Heat management to return prices to $20,
$30 and $40 for the final and deciding Game 7.
In more mainstream consumer transactions, how are reference prices formed? The most common basis for a reference price is the previous price paid for a product. If a consumer has been paying $9.99 for bottle of wine that they have come to enjoy, this $9.99 price becomes this person’s reference price for the wine. If the consumer subsequently encounters the same bottle of wine at the same store for $14.99, the price difference will likely be questioned — not because the wine isn’t worth $14.99 to the consumer, but because the consumer has grown to view $9.99 as the “fair price.”
This concept is captured in our second update to the economic perspective:
Behavioral Update #2: Willingness to Pay is Impacted by Salient Reference Prices
In determining her willingness to pay, a consumer will consider her “economic utility” from the transaction [i.e., perceived value – actual price] and the consistency between the actual price and a salient reference price [i.e., actual price – reference price].
3.
The Impact of a Firm’s Cost of Goods Sold
Scenarios #5 and #6 show that a salient reference price can impact a consumer’s perception of fairness (and, by extension, her willingness to pay). Another factor that impacts a consumer’s perception of “fairness” is a firm’s cost of goods sold, as indicated by the following two scenarios.10
Scenario #7:
A grocery store has no peanut butter in stock, but is about to receive a new shipment. Prior to delivery, the owner finds out that the wholesale price of peanut butter has increased 20% and will affect this new shipment. The owner decides to increase the price of the new peanut butter by 20%. Is this retailer’s actions fair or unfair?
10 These two scenarios are adapted from Kahneman, Knetsch and Thaler, “Fairness as a Constraint on Profit
Seeking: Entitlements in the Market,” American Economic Review 76, 4 (1986): p. 728-741.
8
54
Note on Behavioral Pricing
599-114
With little argument, most consumers find this retailer’s actions to be entirely “fair.” After all, the retailer is only passing along a wholesale price increase to the consumer. The retailer has not caused the price increase and is not benefiting from it. Contrast that with the following scenario:
Scenario #8:
A grocery store has a one week supply of peanut butter in stock and is due to receive a new shipment in the near future. Prior to delivery, the owner finds out that the wholesale price of peanut butter has increased 20% and will affect the new shipment. The owner decides to immediately increase the shelf price on his current stock of peanut butter by 20%. Is this retailer’s actions fair or unfair?
In thinking about these two scenarios, let us again consider the purely economic perspective.
In particular, one’s “perceived value” for peanut butter should not differ between the two scenarios – the peanut butter is the same in both scenarios. In addition, the price being charged for the peanut butter has increased 20% in both cases. Therefore, any sense of “fairness” is not being driven by the
“economic utility” of the transaction [i.e., perceived value – actual price], but by the “psychological utility” of the transaction.
Nonetheless, most consumers find the retailer’s actions “unfair” in Scenario #8. Why? The typical argument is that this retailer is only entitled to pass on a cost increase on a product that has been subject to that increase. In Scenario #8, by raising the price of the in-stock peanut butter that has not been subject to the wholesale price increase, the retailer is “taking advantage of the consumer.”
In addition to the magnitude of their own incentive to purchase, these two scenarios suggest that consumers are concerned with the magnitude of the firm’s incentive to sell [i.e., actual price – cost of goods sold]. In particular, consumers do not want to be taken advantage of and desire a fair division of value between themselves and the firm that makes a product. As a result, they are willing to label a price increase “unfair” when that price increase is coupled with little or no change in the cost of goods sold. This concept is captured in our third update to the economic perspective.
Behavioral Update #3: Willingness to Pay is Impacted by Cost of Goods Sold
In determining his willingness to pay, a consumer will consider his own
“economic utility” from the transaction [i.e., perceived value – actual price] and that of the firm [i.e., actual price – cost of goods sold].
4.
The Nature of the Product Being Sold
To this point, it has been shown that factors such as a salient reference price or a known cost of goods sold can impact perceptions of transaction fairness. How do these perceptions of fairness vary by the type of product being sold? Consider the following scenario:
Scenario #9
In 1996, baseball’s Seattle Mariners made it to the American League playoffs. During the season, general admission to a Mariners game cost $15. For the playoffs, the Mariners raised the price of general admission tickets to $20. Is this fair or unfair?
Most individuals who encounter this scenario feel that this price increase is fair and justified, for many of the same reasons that a ticket price increase in Scenario #5 was fair and justified. In particular, the underlying conditions have changed (playoffs vs. regular season) and consumer
9
55
599-114
Note on Behavioral Pricing
demand almost certainly will increase. Under those conditions, a price increase is reasonable and should be expected. Now contrast this first scenario with the following:
Scenario #10
A hardware store had been selling snow shovels for $15. The morning after a large snowstorm, the store raises the price of its snow shovels to $20. Is this fair or unfair? 11
From an economic perspective, the arguments that support a price increase in Scenario #9 also seem to apply in Scenario #10. In this second scenario, the underlying conditions have changed
(snow vs. no snow) and demand almost certainly will increase.
Nonetheless, while the price increase in the first scenario generally is viewed as “fair,” the price increase in the second scenario is almost universally viewed as “unfair.” Why the difference?
One possible explanation lies in the nature of the product being promoted. Whereas the purchase of tickets to a Seattle Mariners playoff game is viewed as a discretionary expense, a purchase of a snow shovel after a large snowstorm is viewed as a necessary expense. As a result, to raise the price of tickets when your team makes it to the playoffs is viewed as legitimate, but to raise the price of snow shovels after a snowstorm is viewed as exploitation.
These final two scenarios provide our last update to the economic model.
Behavioral Update #4:
Perceptions of Fairness Vary Across Product Categories
In determining her willingness to pay, the degree to which a consumer will rely upon her “economic utility” from the transaction [i.e., perceived value – actual price] will vary across product categories (e.g., discretionary vs. necessity, luxury vs. utilitarian).
Managing Perceptions of Transaction Fairness
This note was designed to highlight some of the psychological drivers of consumer price response. As captured in Figure B, it adds a behavioral component to the more familiar economic approach to product pricing.
Figure B
Combining the Economic and Behavioral Drivers of Willingness to Pay
Consumer
Willingness to Pay
=
Economic Utility of the Transaction
[Perceived Value – Actual Price]
+
Fairness of the
Transaction
[(Perceived Value – Actual Price)/(Actual Price)]
[Actual Price – Expected or Reference Price]
[Actual Price – Cost of Goods Sold]
11 Scenario #10 is adapted from Kahneman, Knetsch and Thaler, “Fairness as a Constraint on Profit Seeking:
Entitlements in the Market,” American Economic Review 76, 4 (1986): p. 728-741.
10
56
Note on Behavioral Pricing
599-114
This framework suggests that an economically rational consumer should decide whether to buy a product solely on the economic utility of the transaction — i.e., by comparing perceived value to product price. In reality, consumers incorporate a host of psychological or behavioral factors into their decision making. These factors include the relative size of the incentive to purchase, the consistency between actual price and expected price, and the difference between product price and the firm’s cost of goods sold. These behavioral factors influence perceptions of transaction “fairness” and tend to reduce a consumer’s willingness to pay relative to a purely economic perspective.
Armed with these insights, however, how can a firm more effectively manage perceptions of transaction fairness and anticipate a consumer’s willingness to pay for its products. Two strategies are recommended.
Strategy #1: Actively Manage Price Expectations
A consumer usually enters a transaction with some expectation about the price of a product.
As evidenced by the scenarios presented in this note, these expectations can have systematic and significant effects on that consumer’s willingness to pay for the product. Unfortunately, firms often do very little to understand and/or manage these expectations.
Instead, firms should actively manage the reference prices and comparisons that consumers employ when assessing product prices. In particular, a firm should look to:
•
Establish credible reference prices.
A firm should look to establish a benchmark price for its products whenever possible. Common tools for doing this include the use of a credible “suggested retail price” or “list price.” Through the clear posting and reliance upon list prices, the automobile industry has done a good job of managing the benchmark price against which consumers evaluate the fairness of the final price paid.
•
Manage product price trends. The single most influential reference price that consumers employ when assessing the fairness of a product price it the previous price paid for that product. As a result, it is far easier to lower prices that are too high than to raise prices that are too low.
•
Encourage favorable comparisons.
Consumers naturally compare the prices of products within and across product categories. In anticipation of this reliance upon comparisons, a firm could suggest comparisons for consumers to consider. For instance, to combat the perception of excessively high prices for its cereals, Kellogg might be welladvised to compare the daily cost of its cereals (about 30¢) to the cost of other breakfast alternatives, such as the cost of a donut or bagel (50¢ or more).
•
Avoid unfavorable comparison through product differentiation. When compact discs first appeared in the music market, they suffered from price comparisons to vinyl records. At $15, CDs were about twice the price of the alternative they were replacing. It was only after convincing consumers that such comparisons were inappropriate, due to vast improvements in sound quality and scratch resistance, that consumers accepted the higher CD prices as fair and reasonable.
Strategy #2: Actively Manage Perceptions of Cost of Goods Sold
As Scenarios #7 and #8 suggest, consumers are sensitive to a firm’s cost of goods sold. As a general rule, consumers are reluctant to pay for products they perceive to be overpriced relative to cost. 11
57
599-114
Note on Behavioral Pricing
This consideration of COGS is especially problematic for firms that operate under high fixed and low variable costs. For example, Microsoft charges hundreds of dollars for its software, yet can churn out incremental copies of that software at the cost of a floppy disk. Many consumers view this as “unfair” and, as a result, have few qualms about using pirated software. In a similar vein, some consumer label drug companies as “greedy” when those companies charge $40 or $50 for a single dose of a branded drug when comparable generics cost a fraction of the price.
Such thinking on the part of the consumer fails to appreciate the fully-loaded cost of product delivery. In the case of Microsoft, the incremental cost of producing another copy of Windows pales in comparison to the cost of developing and supporting that application. And in the case of the drug company, many years and many millions of dollars may have gone into the research and development needed to bring that drug to market. These examples highlight the need for firms to manage the consumers’ perceptions of cost of goods sold. This can be accomplished in several ways.
•
Focus attention of fully-loaded cost of goods sold.
For firms in high fixed cost/low variable cost industries, it is difficult to justify high product prices based on the incremental cost of production. Rather, the firm can focus consumer attention on the fully-loaded cost of production. For instance, if drug companies and software developers effectively communicated the high cost of product development and product support to consumers, they may successfully combat impressions of pricing “unfairly.”
•
Bundle products to obscure cost of goods sold. Some firms sell products where costs are readily apparent to the consumer (e.g., personal computer retailers). Other firms choose to bundle those very same products with additional goods or services so as to obscure the true cost of goods sold. For example, rather than sell computer components, value-added resellers sell turn-key systems. And rather than sell tickets to sporting events, travel agents sell vacation packages that include airfare, hotel and game tickets.
By bundling products, a firm can make its costs less transparent to consumers.
•
Focus attention of consumer value.
A final means by which firms can look to minimize the impact of cost of goods sold is by focusing attention on the consumer’s incentive to purchase. In the end, a consumer benefits whenever perceived value is greater than product price, regardless of the firm’s cost of goods sold. As such, through effective product positioning and communication of value, a firm can minimize the impact of COGS by maximizing attention to the net benefit of purchase to the consumer.
Summary
This note began with the statement, “It is important for a firm to get its pricing ‘right.’” Yet,
“getting prices right” is a complex process that few firms seem well-equipped to manage. While firms have a relatively good grasp of the readily available economic inputs to the pricing decision
(e.g., cost of goods sold, price of substitutes), most lack the behavioral inputs needed to fully understand and anticipate a consumers response to a pricing change. In particular, few firms anticipate the behavioral implications of transaction “fairness,” as perceived by the consumer. This note does not attempt to offer a comprehensive behavioral perspective to pricing. Rather, it is meant to raise awareness that behavioral factors exist and that these factors are important if a firm expects to
“get its pricing right.”
12
58
UVA-M-0675
A PRACTICAL GUIDE TO CONJOINT ANALYSIS
Introduction
Conjoint analysis is a marketing research technique designed to help managers determine the preferences of customers and potential customers. In particular, it seeks to determine how consumers value the different attributes that make up a product and the tradeoffs they are willing to make among the different attributes or features that comprise the product. As such, conjoint analysis is best suited for products that have very tangible attributes that can be easily described or quantified.
While the history of conjoint analysis can be traced to early work in mathematical psychology,1 its popularity has grown tremendously over the last few years as access to easy-touse software has allowed its widespread implementation. There have been probably hundreds of applications of conjoint analysis in industrial settings.2 Some of the more important questions modern conjoint analysis is used to analyze are the following:
1. Predicting the market share of a proposed new product, given the current offerings of competitors. 2. Predicting the impact of a new competitive product on the market share of any given product in the marketplace.
3. Determining consumers’ willingness-to-pay for a proposed new product
4. Quantifying the trade-offs customers or potential customers are willing to make among the various attributes or features that are under consideration in the new product design.
1
R.D. Luce and J. W. Tukey, “Simultaneous Conjoint Measurement: A New Type of Fundamental
Measurement,” Journal of Mathematical Psychology 1 (February 1964): 1–27.
2
P.E. Green, A.M. Kreiger, and Y. Wind, “Thirty Years of Conjoint Analysis: Reflections and Prospects,”
Interfaces 31 (May–June 2001): S56–S73.
This technical note was written by Associate Professor Ronald T. Wilcox. Copyright 2003 by the University of
Virginia Darden School Foundation, Charlottesville, VA. All rights reserved. To order copies, send an e-mail to sales@dardenpublishing.com. No part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the permission of the Darden School Foundation. Rev. 11/03.
59
-2-
UVA-M-0675
The Anatomy of a Conjoint Analysis
Literally, conjoint analysis means an analysis of features considered jointly. The idea is that, while it is difficult for consumers to tell us directly how much each feature of a product is worth to them, we can infer the value of an individual feature of a product by experimentally manipulating the features of a product and observing consumers’ ratings for that product or choices among competing products.
To fix your intuition here, consider the simple example of a sports car. It would be difficult for the average consumer to tell a market researcher exactly how much more valuable a car with 240 horsepower is to them relative to one with 220. It is possible that a consumer might be able to come up with some dollar value, but that value may not really reflect the way they would make choices if faced with a real marketplace situation. Instead, marketers have found that it is much more accurate to present individuals from the target market with a series of cars, described not only by their horsepower but by other attributes as well (color, price, standard/automatic transmission, etc.) and then ask them to rate each of the cars on a numerical scale. Alternatively, the researcher presents several competing cars with different attributes and asks the consumer to choose one. By repeatedly asking the potential customers to rate the cars or chose a car from a competing set, the researcher can infer the value of each individual attribute.
This is the essence of a conjoint analysis; replacing the relatively inaccurate method of asking about each attribute in isolation with a model that allows us to infer the attributes’ values from a series of ratings or choices.
The Experimental Design
A conjoint analysis begins with an experimental design. This design includes all attributes and the values of the attributes that will be tested. Conjoint analysis distinguishes between attributes and what are generally called “levels.” An attribute is self-explanatory. It could be price, color, horsepower, material used for upholstery, or presence of a sunroof, while a level is the specific value or realization of the attribute. For example, the attribute “color” may have the levels “red,” “blue,” and “yellow,” while the attribute “presence of a sunroof” will have levels “yes” and “no.” Before a researcher begins to collect data, it is important that all the levels of each attribute to be tested are written down. Commercially available software packages require that the user provide these as input.
60
UVA-M-0675
-3-
Continuing with our car example, an experimental design might look like the information presented in Table 1:
Table 1. Example of experimental design.
Levels
Price
$23,000
$25,000
$27,000
$29,000
Brand
Toyota
Volkswagen
Saturn
Kia
Horsepower
220 HP
250 HP
280 HP
Upholstery
Cloth
Leather
Sunroof
Yes
No
This is a very simple design that contains a total of 15 attribute levels. Real designs often contain more attributes and levels than are presented here.
When constructing an experimental design, it is important to keep in mind:
1. The more tangible and understandable the levels of each attribute are to the respondents, the more valid the results of the research will be. For example, attribute levels such as “really roomy” are vague, meaning different things to different people and should be avoided.
2. The greater the number of attribute levels to be tested, the more data that will be needed to achieve the same degree of output accuracy.
3. For quantitative variables (price and horsepower in this example), the greater the distance between any two consecutive levels, the harder it will be to get a good idea of how a consumer might evaluate something in between the two (i.e., $24,000).
Data Collection
Collecting data for a conjoint analysis has been made relatively simple by the advent of dedicated off-the-shelf software. The exact nature of the data collected will be dictated by the type of conjoint analysis that is used. An exhaustive discussion of the benefits and drawbacks of each of the many different types of conjoint analysis now in use is beyond the scope of this technical note. However, those interested are encouraged to read Orme for a good discussion of this topic.3
The state-of-the-art in conjoint data collection involves using personal computers or a
Web-based version of the software to guide respondents through an interactive conjoint survey.
The software creates the hypothetical product profiles using the experimental design provided by the researcher and estimates the attribute-level utilities from participant ratings or choices.
3
B. Orme, “Which Conjoint Method Should I Use?” Sawtooth Software Technical Paper (2003). Currently, an
Acrobat-readable copy of this paper is available at http://www.sawtoothsoftware.com/techabs.shtml#which.
61
-4-
UVA-M-0675
Interpreting Conjoint Results
Understanding the basic output
The basic results of a conjoint analysis are the estimated attribute level utilities. Keeping with the example in Table 1, conjoint output might look like Table 2:
Table 2. Conjoint analysis output.
Attribute
Price
Brand
Horsepower
Upholstery
Sunroof
Level
$23,000
$25,000
$27,000
$29,000
Toyota
Volkswagen
Saturn
Kia
220 HP
250 HP
280 HP
Cloth
Leather
Yes
No
Utility (Part-worth)
2.10
1.15
1.56
1.69
0.75
0.65
0.13
1.27
2.24
1.06
1.18
1.60
1.60
0.68
0.68
t-value
14.00
7.67
10.40
11.27
5.00
4.33
0.87
8.47
14.93
7.07
7.87
10.67
10.67
4.53
4.53
The estimated utilities or part-worths correspond to average consumer preferences for the level of any given attribute. Within a given attribute, the estimated utilities are generally scaled in such a way that they add up to zero. So a negative number does not mean that a given level has “negative utility”; it just means that this level is on average less preferred than a level with an estimated utility that is positive.
Conjoint analysis output is also often accompanied by t-values, a standard metric for evaluating statistical significance. Because of the way conjoint utilities are scaled, the standard interpretation of t-values can yield misleading results. For example, the level “Saturn” of the attribute “Brand” has a t-value of 0.87. In general, a t-value of this magnitude would fail a test of statistical significance. However, this t-value is generated because within the attribute “Brand” the level “Saturn” has neither a very high nor very low relative preference. It is basically in the middle in terms of overall preference. Because of the scaling, levels that have more moderate levels of preference within a given attribute are likely to have estimated utilities close to zero, which will tend to produce very low t-values (recall that the t-test is measuring the probability that the true value of a parameter is not different from zero).
62
-5-
UVA-M-0675
A better way to think about statistical significance in this context is to examine the tvalues of the levels with the highest and lowest preference within a given attribute. An applicable common practice would be if the sum of the absolute value of these two statistics is greater than three, then that given attribute is significant in the overall choice process of consumers. At a practical level, it is rare that an attribute will not be significant, and, if you find one that is, it means that it probably should not have been included in the experimental design in the first place, because respondents are not considering that attribute’s information when they make choices. Conjoint Analysis Applications
As mentioned previously, there are many different possible applications of conjoint analysis. We will focus on three very common applications: trade-off analysis, predicting market share, and determining overall attribute importances.
Trade-off analysis
The utility of any given product that we might consider can be easily computed by simply summing the utilities of its attribute levels. For example, a Toyota with 280 horsepower, leather interior, no sunroof, and a price of $23,000 has a utility of 0.75 + 1.18 + 1.60 0.68 + 2.10 =
4.95. If the car with same basic specifications were a Volkswagen, the overall utility would drop to 0.65 + 1.18 + 1.60 0.68 + 2.10 = 4.85, a drop of 0.10. This drop can be seen directly by noticing that the difference between the utility for the brand Toyota (0.75) and Volkswagen
(0.65) is 0.10 and because nothing else in the profile of the car has changed this will be the exact utility difference between two cars that are the same except for this brand difference.
A natural consequence of the above observation is that we use the utilities to analyze what average consumers would be willing to give up on one particular attribute to gain improvements in another. For example, how much money would they be willing to give up
(price) if a sunroof was added to the vehicle? We will now look directly at this issue of the hypothetical car detailed in the previous paragraph. Adding a sunroof to the (Toyota) car would yield an overall utility of 0.75 + 1.18 + 1.60 + 0.68 + 2.10 = 6.31. This represents an increase in utility of 6.31 – 4.95 = 1.36 over the identical car without a sunroof.
The information above directly implies that we can reduce the utility of price by 1.36, and average consumers would be just as happy as before the sunroof was installed. To find out how much the price can be raised, we must convert the change in utility with a change in price.
We do this by first noting how much the original car costs ($23,000) and the utility associated with that figure, 2.10. We know that we can reduce the price utility by 1.36. This is equivalent to saying that we can reduce the price utility to 2.10 – 1.36 = 0.74. By referring to Table 2, we can immediately see that this implies a price between $25,000 and $27,000 because 1.56 < 0.74 <
1.15. In fact, if we assume a linear relationship between price and utility in the range between
63
UVA-M-0675
-6-
$25,000 and $27,000, we can solve for the exact price by performing a linear interpolation within this range.4 Specifically, the interpolation yields:
$25,000
1.15 0.74
$2,000
1.15 ( 1.56)
Utility spread between the two tested price points (25K and 27K).
$25,302.58
Utility spread between 25K and the target utility.
This implies that, if the sunroof is added, the price of the vehicle could be raised from
$23,000 to about $25,300, and the average consumer’s attitude would be one of indifference between the two vehicles. Qualitatively, it shows that the value of a sunroof to consumers is very substantial. This same kind of analysis can be performed for other attributes. We could ask how much additional horsepower we would need to add if the interior was changed from leather to cloth.
This particular question does present a problem, however. Because the current vehicle under consideration has 280 horsepower, and that is the maximum amount of horsepower tested by the conjoint analysis, it will be impossible to determine how much consumers will value additional horsepower. This leads to an important consideration when constructing the experimental design.
That is, if the output is to be used for trade-off analysis, it is important that the range of the levels tested within each attribute span the entire range of that attribute before management would ever consider it as a realistic design alternative. If the experimental design takes this into account, we can perform trade-off analysis between any two attributes in the design.
Market share forecasting
Another common application is forecasting market share. In order to use conjoint output for this kind of prediction, two conditions must be satisfied:
1. The company must know the other products, besides their own offering, that a consumer is likely to consider when making a selection in the category.
2. Each of these competitive products’ important features must be included in the experimental design. In other words, you must be able to calculate the utility of not only your own product offering but that of the competitive products as well.
4
This is a common way to approximate the relationship between the value of the attribute and its utility for attribute values that were not directly tested by the conjoint analysis. The closer the tested levels are to each other the more accurate this approximation. Also notice that this interpolation can only be performed for quantitative attributes such as price. Interpolating between qualitative attributes, like brand, is nonsensical.
64
UVA-M-0675
-7-
Market share prediction relies on the use of a multinomial logit model.5 The basic form of the logit model is: eU i
Sharei
n j 1
Uj
e
where:
Ui is the estimated utility of product i
Uj is the estimated utility of product j n is the total number of products in the competitive set, including product i.
To make things clear, consider the following example. Suppose we are interested in predicting the market share of a car with the following profile: Saturn; $23,000; 220 HP; cloth interior; no sunroof. We believe that when consumers consider our car they will also consider purchasing cars that are currently on the market with the following profiles:
1. $27,000; Toyota; 250 HP; cloth interior; no sunroof
2. $29,000; Volkswagen; 280 HP; leather interior; no sunroof
3. $23,000; Kia; 220 HP; cloth interior; no sunroof
For the Saturn and its associated product profile the estimated utility is 2.10 – 0.13 – 2.24
– 1.60 – 0.68 = 2.55. Similarly, the utilities of the three competing products can be calculated:
1.
2.
3.
1.56 + 0.75 + 1.06 – 1.60 – 0.68 = 2.03
1.69 + 0.65 + 1.18 + 1.60 – 0.68 = 1.06
2.10 – 1.27 – 2.24 – 1.60 – 0.68 = 3.69
With these utilities in hand we can now directly apply the logit model to forecast market share for the Saturn. This is given by:
ShareSaturn
e e 2.55
e
2.03
2.55
e1.06
e
3.69
.025 or 2.5%
This implies that this particular Saturn vehicle will achieve a 2.5% market share within the specified competitive set. The market share of any vehicle that can be described by the experimental design and a set of competitive vehicles, also described by the experimental design, can be found in a similar manner.
5
A good marketing reference to learn about the basics of the logit model is G. Lilien and A. Rangaswamy,
Marketing Engineering: Computer-Assisted Marketing Analysis and Planning (2nd Ed.), Englewood Cliffs:
Prentice-Hall (2002). Also, most econometrics textbooks will have information on logit models.
65
UVA-M-0675
-8Determining attribute importance
A researcher may also be interested in determining the importance of any individual attribute in the consumers’ decision processes. Quantifying these attribute importances using the conjoint output is straightforward and can provide both interesting and useful insights into consumer behavior.
Intuitively, the variance of the estimated utilities within a given attribute tells you something about how important the attribute is in the choice process. Take, for example, the attributes “Sunroof” and “Upholstery,” both of which have only two levels. If you understand the material up to this point, it should be reasonably clear that “Upholstery” is a more important attribute than “Sunroof.” That is because the utility difference between having a sunroof and not having a sunroof (2 × 0.68 = 1.36) is smaller than the utility difference between having leather versus cloth interior (2 × 1.60 = 3.20).
The common metric used to measure attribute importances is:
Ii
Ui n j 1
Ui
Uj
U
j
where:
Ii = the importance of any given attribute i
U = the highest utility level within a given attribute (subscripts indicate which attribute)
U = the lowest utility level within a given attribute.
This equation is really quite intuitive. In order to calculate the importance of any given attribute, you just take the difference between the highest and lowest utility level of that attribute and divide this by the sum of the differences between the highest and lowest utility level for all attributes (including the one in question). The resulting number will always lie between zero and one and is generally interpreted as the percent decision weight of an attribute in the overall choice process.
It also should be clear at this point that this estimated attribute importance depends critically on your experimental design. In particular, if you increase the distance between the most extreme levels of any given attribute you will almost certainly increase the overall attribute importance. For example, if the tested price range was $21K $31K instead of $23K – $29K
(Table 1), this is very likely to increase the estimate attribute importance of price.
Let’s now consider a concrete example using the attribute “Horsepower.” The importance of this attribute is calculated as:
66
-9-
I Horsepower
UVA-M-0675
1.18 2.24
= .25
(2.10 1.69) (0.75 1.27) (1.18 2.24) (1.60 1.60) (0.68 0.68)
In our example, 25% of the overall decision weight is assigned to horsepower. The reader may verify through analogous calculations that the decision weight for “Price” is about 27%,
“Brand” about 15%, “Sunroof” about 10%, and “Upholstery” about 23%. The numbers provide a very intuitive metric for thinking about the importance of each attribute in the decision process.
Final Thoughts
Conjoint analysis has a broad array of possible applications. Many of these applications are variants of the three very common applications presented here. The increasingly widespread availability of conjoint analysis software, both PC and Web-enabled, points to its continued growth as a marketing decision aid.
This note has presented what is generally known as “aggregate-level” conjoint analysis.
That is, all of the respondents are pooled into one group, and a single set of attribute level utilities are estimated from the ratings or choices provided by the people in this group. Recent advancements in conjoint analysis have enabled researchers to estimate different utilities for different groups of respondents and even, in some cases, for individual respondents. While the mathematics necessary for this procedure is sometimes quite complex, it is now possible to estimate the attribute level utilities and to compute trade-off analyses for each individual respondent. This has some significant advantages over aggregate level analysis particularly when considering marketing segmentation issues. Either way, the data collection and the basic interpretation of the output remains the same. While there is currently no textbook that can provide the reader with answers to all of the questions that might arise when applying this technique in a business setting, there is, as of this writing, a very good and surprisingly comprehensive collection of technical papers located on the site of a company that markets conjoint analysis software (http://www.sawtoothsoftware.com/techpap.shtml). These provide answers to many of the practical implementation questions a user may face.
67
68
UVA-M-0674
FIDELITY INCORPORATED:
PRICING THE FIDELITY BLUE CHIP GROWTH FUND
John McDowell walked off of the 18th green at Andover Country Club and handed his putter to his caddie in obvious disgust. He had double-bogeyed the 18th and, in doing so, shot an
84, his worst round of golf in the last several months.
McDowell knew what the problem was. It was not his stance, his shoulder position, or even his often-beguiling driver. It was his lack of concentration brought on by the pricing decision facing him, a decision upon which he knew millions of dollars were at stake.
As the fund manager of Fidelity’s Blue Chip Growth Fund (BCGF), McDowell not only oversaw the investment decisions of the fund but had considerable input into its marketing plan as well. Senior management at Fidelity had asked McDowell to consider how to generate more profit from the fund. While the fund had grown significantly over the last few years, management believed that they were not charging customers enough for the services provided.
Fidelity had recently spent a great deal of money updating and enhancing their on-line order handling and customer service capabilities and believed it was now appropriate to ask their customers to pay something for these increased services.
BCGF currently had fees that were significantly below the industry averages for actively managed domestic equity mutual funds. Investors in this fund paid no front- or back-end load and annual expenses of 77 basis points (.77 percent).1 Total expenses attributed to the fund and expenses for fund management and marketing came to about 45 basis points. As of May 2003,
BCGF held $17.9 billion in assets.
Most BCGF investors viewed management of their investments as a reasonably long-run decision. While the average holding time for an investment in the fund fluctuated with economic and market conditions, management believed the average holding time to be around seven years.
This number was typically larger for investors whose money was tied to a 401(k) or other tax1
A front-end load is a one-time charge, generally expressed as a percent of the assets invested (i.e., 2 percent).
A back-end load is analogous to a front-end load except the fee is paid when fund shares are redeemed. Annual expenses, also expressed as a percent of assets under management, are deducted from the fund on an ongoing basis.
This case was prepared by Ronald T. Wilcox, associate professor of Business Administration, Darden Graduate
School of Business Administration, University of Virginia. It was written as a basis for class discussion rather than to illustrate effective or ineffective handling of an administrative situation. Copyright 2003 by the University of
Virginia Darden School Foundation, Charlottesville, VA. All rights reserved. To order copies, send an e-mail to sales@dardenpublishing.com. No part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the permission of the Darden School Foundation.
69
-2-
UVA-M-0674
defined contribution pension plan and typically less for those holding the funds in a regular, nontax-advantaged, brokerage account. In the previous year about 15 percent of the mutual fund shares held by investors at the beginning of the year were sold, and the money moved elsewhere.
The fund also received new monies in the amount of about 20 percent of the value of the fund.
Both of these figures represent dollars-in-the-door and dollars-out-of-the-door based on customers’ decisions and are not related to any investment gains or losses experienced by the fund. So, the fund experienced a net cash inflow of 5 percent during the previous calendar year.
While these types of inflows and outflows were difficult to predict, McDowell believed a similar pattern might hold in the coming year. Table 3 provides some information on the growth of mutual fund ownership among U.S. households.
Edward Johnson, CEO of Fidelity Investments, in a memo to McDowell earlier in the week, had asked him to explore ways to increase fund profitability by 30 percent. It was this task that had McDowell concerned. He didn’t know whether he should try to generate the additional profit by introducing a front-end load to the fund or by increasing the annual fees.
Back-end loads were probably not a feasible alternative given the history and pricing philosophy of Fidelity. And, although any front-end load was technically possible, industry norms dictated that loads were set in increments of 50 basis points. So, a load of 1.5 percent was fine in terms of industry norms, but a load of 0.75 percent was not.
McDowell was concerned that potential customers might view the higher fees as a deterrent to investing in BCGF, and worried that even current customers would leave the fund.
He knew he needed to find a way to generate the additional profit in a way that would lead to the least dissatisfaction among current and potential future investors.
McDowell’s assistant, Maria Ramos, had pointed out a recently published study that might help him make the new pricing decision. The study used the marketing research technique, conjoint analysis, to examine investors’ preference for stock mutual funds. This particular conjoint study had estimated the utilities or preferences of investors for different levels of front-end loads and annual expenses among other fund attributes. Ramos pointed out that the results of this study might shed light on how to increase profitability without seriously harming present or future investor satisfaction with the fund.
The Conjoint Study
A detailed description of the conjoint analysis of mutual fund choice was published in a technical academic journal.2 Ramos distilled what she believed to be the important facts for the decision at hand and brought them to McDowell. These facts included the important features of the experimental design, given in Table 1 and the results of the conjoint estimation, given in
Table 2. The headings of the columns in Table 1 indicate the name of the tested attribute and the values below the heading the levels of each of the tested attributes. Table 2 provides the estimated utilities or part-worths for each level of each attribute.
2
R.T. Wilcox, “Bargain Hunting or Star Gazing? Investors’ Preferences for Stock Mutual Funds,” Journal of
Business 76 (2003).
70
-3-
UVA-M-0674
Table 1: Experimental Design
Front-End Load
Annual Fee
10-Year Past Return
Beta3
No Load
0%
5%
0.7
1%
0.5%
10%
0.9
2%
1.0%
15%
1.1
3%
1.5%
20%
1.3
4%
2.0%
25%
1.5
5%
2.5%
________________________________________________________________________
Company
Fidelity
Vanguard
T. Rowe Price
Dreyfus
Pecunia
The study appeared to be conducted using a representative sample of mutual fund investors, but not necessarily customers of Fidelity BCGF or any other Fidelity mutual fund.
Ramos reported to McDowell that the data was collected using a specific type of conjoint called
Choice-based Conjoint analysis, a type that seemed appropriate for this application.4 Among the other results of the study, the author had also concluded that investors appeared not to take into account their investment time horizon when evaluating the fees of a given fund.
In addition to the question of price, McDowell also believed that Fidelity’s brand name was very strong and, based on this brand name, he could probably charge a premium to his chief competitor, Vanguard. Vanguard had been aggressively marketing their PRIMECAP fund, a fund with a very similar asset base, to BCGF. PRIMECAP charged no front-end load and had an annual fee of 49 basis points. McDowell believed that Fidelity’s superior customer service, investment in advertising, and enhanced Web interface sufficiently increased investors’ perception of the Fidelity brand to insulate BCGF from any serious price competition from
Vanguard.
The Pricing Decision
The stakes for this particular pricing decision were huge. McDowell, scribbling some quick calculations on a napkin, realized that every additional basis point they were able to capture in profit would increase fund profitability by almost $2 million. He struggled with how to use the information from the conjoint analysis to help his pricing decision. He wondered whether a study published in some academic journal he had never heard of, using a sample of investors who were not necessarily his customers, was useful at all in this situation. He felt confident that the Fidelity brand would keep him from losing investors even in the face of the price increase. Ultimately, he told himself, the increased profit target of 30 percent is not my call; my job is to figure out how to get this done.
3
Beta is a measure of the covariance of the return of a fund relative to an index, in this case the S&P 500. It is often used as a measure of the volatility of a fund. For example, a beta of 1.3 indicates that the fund is 1.3 times more volatile than the S&P 500.
4
B. Orme, “Which Conjoint Method Should I Use?” Sawtooth Software Technical Papers (2003).
71
UVA-M-0674
-4Table 2: Results of Conjoint Analysis
Utility Estimate
St. Dev. of Estimate
Fidelity
Vanguard
T. Rowe Price
Dreyfus
Pecunia
0.08
0.01
0.08
0.00
-0.17
0.09
0.09
0.09
0.09
0.09
No Load
1%
2%
3%
4%
5%
0.88
0.48
0.05
-0.25
-0.49
-0.67
0.10
0.10
0.11
0.11
0.13
0.13
0%
0.5%
1.0%
1.5%
2.0%
2.5%
0.99
0.50
0.05
-0.07
-0.72
-0.74
0.10
0.10
0.11
0.11
0.13
0.13
5%
10%
15%
20%
25%
-1.23
-0.70
-0.06
0.76
1.23
0.10
0.10
0.10
0.10
0.10
0.7
0.9
1.1
1.3
1.5
0.05
0.06
0.13
-0.10
-0.14
0.10
0.10
0.10
0.10
0.10
Company
Load
Annual Expenses
10-Year Return
Beta
72
UVA-M-0674
-5Table 3: The Growth in Mutual Fund Ownership
U.S. Households Ow ning Mutual Funds, 1980-2001
52.0%
49.0%
44.0%
37.2%
30.7%
24.4%
25.0%
27.0%
20.0%
11.0%
11.9%
5.7%
1980 1982
1980
4.6
1982
9.0
1984
1984
10.2
1986 1988
1986
17.3
1988
22.2
1990 1992
1990
23.4
1992
25.8
1994 1996
1994
30.2
1998
1996
36.8
2000 2001
1998
44.4
2000
51.7
Millions of U.S. Households
Source: Investment Company Institute, “Fundamentals: U.S. Household Ownership on Mutual Funds in 2001.”
73
2001
54.8
74
UVA-M-0773
Rev. Feb. 22, 2010
PORTLAND TRAIL BLAZERS
“We have a new group sales operation in place and we’re looking for dramatic results there.”
—Steve Patterson, President
Portland Trail Blazers
Less than a month after the 2005 NBA All-Star break, the Portland Trail Blazers was a team in upheaval. On the court, they had just fired their coach of the past four seasons and were
22–36, in danger of one of the worst seasons in franchise history. Off the court, the Blazers organization was facing considerable challenges as well. The team’s home arena, the Rose
Garden, had filed for Chapter 11 bankruptcy and was being run by the building’s creditors.
The arena, a virtual lock to sell out just three seasons ago, had seen attendance numbers fall more than 15% since the 2003 season (Figure 1). During the same time, the organization had only been successful in renewing 9 of the 46 luxury-suite contracts that came due in 2005, and
42 of the 70 luxury suites sat empty during the season.1 Television interest also declined, with a
Portland-area Nielsen share of just 5% when the Blazers played the Minneapolis Timberwolves
(weather coverage generally received up to 20%).2
1
2
Todd Murphy, Todd. “Have Arena, Need People.” Portlandtribune.com, August 10, 2004.
Pete Schulberg, “Blazers Start Losing with Viewers, Too,” Portland Tribune, January 21, 2005.
This case was prepared by Professor Ron Wilcox. It was written as a basis for class discussion rather than to illustrate effective or ineffective handling of an administrative situation. Copyright ¤ 2009 by the University of
Virginia Darden School Foundation, Charlottesville, VA. All rights reserved. To order copies, send an e-mail to sales@dardenbusinesspublishing.com. No part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the permission of the Darden School Foundation. Rev. 2/10. ¸
75
-2-
UVA-M-0773
Figure 1. Average home attendance, 1996–2007.
Attendance
21,000
20,000
19,000
15%
18,000
17,000
16,000
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Season
Source: Portland Trail Blazers.
A similar story was occurring in the sale of the team’s “club seats”—special seats for
Blazers games that were sold on a multiyear contract and came with club perks. Of 1,800 club seats in 2005, 700 remained available, most of them because subscribers had dropped their contracts during the previous season.
The Portland Sports Market
The Trail Blazers had a monopoly on the professional sports market in Portland, Oregon.
Without a dominant university sports program affiliated with the city, the team competed only with minor league baseball and hockey for its share of the city’s sporting dollars. At just under two million people in the metro area, Portland was the fourth-smallest market in the NBA.
Each year the Blazers tracked Portland-area residents’ general perception of the team (see
Figure 2). The historically strong relationship between the team and the city had soured over the past few seasons, with the percentage of people perceiving the team negatively having increased tenfold since 2000. Fan support had dwindled due to a number of widely publicized player transgressions including marijuana use, fighting with teammates, and an incident involving animal cruelty.3
3
Andy Giegerich, “Beleaguered Blazers Play by the Numbers,” Portland Business Journal, October 29, 2004.
76
-3-
UVA-M-0773
Figure 2. Portland-area residents’ general perception of the Portland Trail Blazers.
100%
92%
Favorable
Unfavorable
80%
71%
60%
40%
Jun-01
Dec-01
60% 63%
56%
53%
51%
42%
47%
49%
29%
24%
20%
5%
0%
Jun-00 Dec-00
68%
Jun-02
Dec-02
Jun-03
54%
46%
40% 37%
Dec-03
Jun-04
57%
43%
Dec-04
Source: Portland Trail Blazers.
Multigame Ticket Packages
One of the more successful Blazers promotions during the past few seasons had been multigame ticket packages. This program allowed fans to purchase tickets for a number of games at once, and usually included at least one marquee opponent, a game for which individual tickets were difficult to find. Trail Blazers’ management saw this program as an effective tool to:
1. Increase ticket sales for less popular games (typically bundled in the package with tickets for hard-to-find games)
2. Increase overall ticket sales because the multigame packages acted as an effective incentive for those who planned on attending only one or two games during the season to increase the number of games they attended
3. Develop more of an ongoing relationship with fans who could potentially become future season ticket holders
Despite the program’s relative success, management wanted to explore all potential packages, and better understand which options were most popular with fans. The program’s goal was to offer a multigame ticket package that had a high appeal to fans, while being profitable to the team and not undermining current pricing policies.
77
-4-
UVA-M-0773
Designing the Research Study
The Trail Blazers management team hired Acuity Market Research, a Portland-based research firm, to help design their multigame package study. Together, they determined there were six aspects of the multigame ticket packages that drove a customer’s decision to purchase:
1. The team the Blazers were playing
2. The day of the week the game was being played
3. The number of games included in the package
4. The location of the seats
5. The price (per seat) of the package
6. The promotional item included in the package
The project team designed a study utilizing conjoint analysis to ascertain the importance of the individual attributes, as well as the likely response of the market to specific multigame ticket packages. Some things were givens: (1) there was a high number of teams in the NBA (30 including the Blazers) and (2) the dates of the games included in the package could not be changed, those attributes would not be included as part of the conjoint products. Instead, questions pertaining to favorite teams and days to watch a game were asked individually, after the conjoint portion of the survey.
An e-mail went out from the Blazers’ director of database and Internet marketing to 960 fans who had purchased multiple-game ticket packages or season tickets in the past, but were not current season ticket holders. The project team decided it was more important to get feedback from people who had expressed some level of commitment to purchasing Blazers tickets in the past than general fans of the team, and knew it had current e-mail addresses for this group.
Although new fans were always purchasing multigame or season ticket packages, the Blazers’ management believed past purchasers were likely the best prospects for new multigame packages. The initial e-mail explained the purpose of the study and asked fans to participate. One week later, a reminder e-mail was sent in hope of increasing the overall response rate of the study. Both e-mails contained a link to an online conjoint-based survey which included 20 different conjoint choice tasks, (an example is included in the Appendix) Blazers-specific questions, and a battery of demographic questions. Most respondents took between 10 and 15 minutes to complete the survey.
Participants were given a chance to win free tickets to Blazers games, as well as autographed Blazers items such as jerseys, basketballs, and posters for taking part in the survey.
78
-5-
UVA-M-0773
Study Findings
The e-mail solicitations received a total of 204 valid responses (a 21% response rate).
Summary statistics regarding demographics and past Blazers-game attendance are located in the
Appendix.
Acuity began their analysis of the multigame packages by computing the attribute level utility scores to help better understand the stand-alone preference of each of the individual attribute levels. The utility score data are shown in Table 1.
Table 1. Utility score data.
Utility
0.03257
0.24383
–0.2764
Utility
0.65646
0.22011
0.126
–1.00257
Utility
–0.73169
–0.43716
0.15736
1.01148
Utility
0.12511
0.17428
0.00158
–0.31786
0.01689
Number of Games
3-game create-your-own pack, including 1 elite team and 2 very good teams
6-game create-your-own pack, including 2 elite teams and 4 very good teams
10 game create-your-own pack, including any combination of teams.
Ticket Price
$15 per seat per game
$25 per seat per game
$35 per seat per game
$60 per seat per game
Ticket Location
300 level, behind the baskets
300 level, on the corners
300 level, midcourt
200 level, midcourt
Promotional Item
Priority for home playoff tickets
Hot dog and soda with each ticket
Trail Blazer apparel (hat, jersey, etc.)
Trail Blazer collectible (bobblehead, etc.)
$20 gift certificate for popular local restaurant
Source: Portland Trail Blazers.
While the conjoint study allowed all the attributes and levels to be randomly assigned, in reality the Blazers were unwilling to allow certain price and seating combinations no matter how well received they were, due to the cost structure of the arena. They disallowed: x 200-level seats for less than $60
x
300-level, midcourt seats for less than $25
79
-6-
UVA-M-0773
Cost of Multigame Packages
While the fan preference was extremely important to Blazers’ management, any multigame packages the group designed had to be financially attractive and align with the organization’s strategic goals. Each of the multigame ticket package attributes had costs and strategic implications associated with them:
Number of Games: The Blazers prefer the six-game package because it offered the capability of pairing the most popular teams with games that were more difficult to sell tickets to
(weekday games, less competitive teams, etc.). Their next preference would be a 10-game package, because it allowed the team to efficiently sell a large number of the remaining games to a single fan.
Seat Location: Although nearly all the Blazers’ stadium costs were fixed expenses, the organization still applied a cost to each of the seat locations in the stadium. This cost structure had to be met, at a minimum, for any tickets that were sold and differed based on seat location.
Minimum seat pricing is shown in Table 2.
Table 2. Fixed costs based on seat location.
Seat Location
Fixed Cost
300 level, behind the baskets
$10.00
300 level, on the corners
$12.00
300 level, midcourt
$18.00
200 level, midcourt
$40.00
Source: Portland Trail Blazers.
Promotional Items: A direct cost is associated with each of the promotional items the
Blazers may offer fans. For example, if the Blazers were to offer a hot dog and soda with each ticket, they would have to pay the Rose Garden’s vendor services a negotiated price of $3.25 per package. The $20 gift certificate to a popular restaurant is purchased for a negotiated price of
$10. The restaurateur deeply discounts the gift certificates in exchange for the marketing exposure. The only promotional item without a direct cost was offering priority for home playoff tickets, given that the tickets are still sold at full retail price and multigame ticket holders just receive priority in purchasing available tickets. Table 3 presents the unit cost of each of the potential promotional items.
80
-7-
UVA-M-0773
Table 3. Cost of promotional items.
Promotional Item
Cost
Priority for playoff tickets
$0.00
Hot dog and soda with each ticket
$3.25
Trail Blazers apparel (hat, jersey, etc.)
$12.00
Trail Blazers collectible (bobble head, etc.) $6.75
$20 gift certificate to a popular restaurant
$10.00
Source: Portland Trail Blazers.
Utilizing the conjoint information, in addition to the other data available from the survey, the Blazers’ management team felt prepared to design the multigame package they believed the fans would most prefer. What attributes were most important to the fans? What should the
Blazers’ multigame package include? Should there be more than one? How profitable would each of the packages be?
81
-8Appendix
PORTLAND TRAIL BLAZERS
Example: Online Conjoint Survey
Source: Portland Trail Blazers.
82
UVA-M-0773
-9-
UVA-M-0773
Appendix (continued)
Study Demographics
People under the age of 18 in the home
Gender
Female,
21.9%
57.5%
Zero
Male,
78.1%
22.5%
One
15.5%
Two
3.5%
Three
1.0%
Four or more
0
0.1
0.2
0.3
0.4
0.5
10.4%
18-24
65.2%
Married
25-34
27.8%
26.9%
30.8%
35-44
20.4%
45-54
6.1%
Divorced
55-64
1.0%
Widowed
0
6.5%
5.0%
65 and over
0.1
0.2
0.3
0.4
0.5
0.6
0.0%
0.7
5.0%
Under $20,000
3.5%
Some College (including vocational and technical school)
26.4%
0.2
0.3
6.3%
$30,000 to $39,999
6.3%
25.0%
30.0%
35.0%
7.9%
9.4%
$60,000 to $69,999
11.5%
25.1%
$75,000 to $99,999
20.4%
0.1
20.0%
2.6%
$50,000 to $59,999
49.8%
0
15.0%
$20,000 to $29,999
$40,000 to $49,999
College Graduate
Advanced Degree
10.0%
Annual Household Income
Educational Level of Respondents
High School Graduate
0.7
Age of Respondents
Marital Status of Respondents
Single
0.6
30.9%
$100,000 or more
0.4
0.5
0.6
0
Source: Portland Trail Blazers.
83
0.05
0.1
0.15
0.2
0.25
0.3
0.35
-10-
UVA-M-0773
Appendix (continued)
Game Attendance Behavior
Have purchased multi-game tickets
Have purchased single-game tickets
No, 23.2%
No, 41.9%
Yes, 58.1%
Yes, 76.8%
Have purchased season tickets
Have purchased other tickets
Yes, 4.0%
Yes, 15.2%
No, 84.8%
No, 96.0%
Number of Trail Blazer home games attended in past two seasons
None
2.7%
1-2
3-5
5.9%
20.0%
6-10
11-15
13.5%
16-20
21-30
31-40
More than 40
0.0%
23.2%
20.5%
3.8%
8.6%
1.6%
5.0%
10.0%
Source: Portland Trail Blazers.
84
15.0%
20.0%
25.0%
85
Source: Portland Trail Blazers.
Rose Garden Seating Chart
Appendix (continued)
-11-
UVA-M-0773
86
9-503-019
AUGUST 7, 2002
ELIE OFEK
Customer Profitability and Lifetime Value
The digital revolution has given rise to a host of technologies that are transforming marketing practices. Powerful databases and electronic data networks are allowing companies to collect concise information about customers and their buying patterns more effectively and efficiently than ever before. The Internet, in particular, has increased the ability of firms to track the behavior of individual consumers as they visit numerous Web pages.
While these capabilities challenge companies to manage vast amounts of information, they offer marketers exciting new opportunities to dynamically manage their customer bases. Beyond the traditional focus on mass advertising, firms can now more accurately recognize and monitor individual customers to whom tailored communications and offers can be made. Marketing activity is becoming more interactive and engaging. For many companies, particularly large or multinational corporations, this was altogether infeasible or prohibitively costly in the past.
In order to reap the benefits of detailed customer knowledge, firms need to systematically estimate the profitability associated with its use. The ultimate goal is to develop highly committed customers who not only make repeat purchases and generate continual revenue streams, but also require minimal maintenance along the way. It is entirely possible that while some customers do not bring profits with their initial purchases, the margins from their future expected transactions paint a different picture. As a result, firms need to track initial acquisition costs and compare them to the profits to be generated over the customer’s expected relationship with the company. The above activities allow marketers to decide which customers to go after, how to change the promotional mix as a function of past and recent transactions and, if necessary, which specific customers to discontinue serving. Indeed, many practitioners and scholars are expressing the view that marketing is rapidly becoming “the science and art of finding, retaining, and growing profitable customers.”1
The initial step in determining customer profitability is to have a clear sense of the relevant characteristics of customer activity. To this end, analysis of historical data can be very powerful in providing firms with information on the buying patterns of both new and existing customers (these may or may not be correlated with certain demographics). Using this information, the firm may choose to distinguish between segments of customers that exhibit similar characteristics and are hence expected to respond similarly to a given level of communication/promotion. Taking into account the responsiveness of particular customers to certain offerings, the cost of making these offerings, and the probability a customer is expected to keep generating revenue for the firm enables
1 Philip Kotler and Gary Armstrong, Principles of Marketing (Upper Saddle River, NJ: Prentice Hall, 2001).
________________________________________________________________________________________________________________
Professor Elie Ofek prepared this note with the assistance of Research Associate Kevin Morris as the basis for class discussion rather than to illustrate either effective or ineffective handling of an administrative situation.
Copyright © 2002 President and Fellows of Harvard College. To order copies or request permission to reproduce materials, call 1-800-545-7685, write Harvard Business School Publishing, Boston, MA 02163, or go to http://www.hbsp.harvard.edu. No part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the permission of Harvard Business School.
87
503-019
Customer Profitability and Lifetime Value
the calculation of customer profitability and lifetime value. These ideas are made clearer through the following example.
Example: The Catalog Retailer
Part I—Customer Acquisition Costs
A direct catalog retailer of fashion goods is contemplating whether or not to attract new customers using names purchased from a list broker or by randomly sending out catalogs. The cost of sending a catalog (which includes production and mailing) is $0.5. From experience, the company anticipates the response rate (the percentage of individuals who receive the catalog and purchase from it, hence becoming “current customers”) from a random mailing to be 1%. By analyzing the buying behavior and demographics of current customers, the retailer estimates it can rent names selectively from the broker to achieve a response rate of 4%. The list broker has set a price of $0.2 for each name. The direct retailer wishes to calculate the acquisition costs associated with each source of potential customers. To determine acquisition costs, the retailer needs to use the response rate and cost of sending a catalog to each prospect.2 A response rate of 1% (or 0.01) means that of 100 catalogs sent, only one recipient is expected to respond. Similarly, a response rate of 4% means that out of 25 catalogs sent, one recipient is expected to respond with a purchase. The cost of acquiring a customer can then be calculated using the following:
Acquisition Cost= (# of catalogs needed to get 1 customer) * (cost of sending a catalog) =
cost of sending a catalog response rate
Thus, the acquisition cost using the random mailing approach is: 0.5/.01=$50, and the cost of acquiring a customer from the rented list is: 0.7/0.04= $17.5. The retailer concludes that even though each rented name increases the effective cost of sending out each catalog by 40%, the improvement in targetability makes this a worthwhile avenue to pursue.
Additional Comments:
a) In some instances, the response rate is not given directly but can be calculated using other pieces of information. For example, direct marketers often refer to subsets of customers by the “demand per book” they generate. Demand per book is the expected revenue from one mailed catalog. If the average order size (in dollars) is known, the response rate can be calculated by dividing these two quantities, that is, response rate = (demand per book)/(average order size).
b) As any rented list of names includes a finite number of prospects that meet certain criteria, the retailer must also take into account that at some point all relevant names will be exhausted. At this point, to acquire new customers, the retailer can either move on to names expected to yield lower rates of return (but that are still profitable), or attempt to find new sources of names that were not on the previous broker’s list.
2 The marketing effort to acquire new customers from a pool of potential individuals is often called “prospecting.”
2
88
Customer Profitability and Lifetime Value
c)
503-019
The above calculation of acquisition costs can be generalized to other industry contexts. To do so, one needs to recognize the instruments used to generate a response (such as a visit by a salesperson, a call from a telemarketer, mailing of an informational brochure, free samples, or special discount coupons) and the likelihood of obtaining a response.
Part II—Customer Break-Even Analysis
Now that we have determined the cost of acquiring a customer, we can use this information to establish how many purchases or years it will take for each customer to realize profits for the firm.
In our example, the catalog retailer makes changes to the product line four times a year. Historical data has shown that one can roughly segment the market of current customers after one year into frequent buyers, who purchase twice a year with an average order size of $50, and occasional buyers, who purchase only once a year with an average order size of $80. The retention rate (percentage of customers who continue to make a purchase with the company in the next period) is 75% for frequent buyers and 50% for occasional buyers. Thus, the likelihood an individual acquired today will remain a customer by year five, or the survival rate of year five, is approximately 30% for the frequent buyers and 6% for the occasional buyers.3 Gross margins are 20% of sales, and include all expenses aside from the cost of sending out catalogs. In the first year, the retailer sends a catalog every month (with the same catalog sent in three consecutive months) to all acquired customers. Based on first year purchase patterns, frequent buyers continue to receive 12 catalogs a year, while occasional buyers receive only four. The catalog retailer is interested in knowing how many years it would take to recoup initial costs of acquisition, assuming customers were acquired using rented names.
To address the retailer’s question, it is useful to construct the following table for each of the two segments: Figure A
Customer Break-Even Analysis for Frequent and Occasional Buyers
In both tables below: In line D, expected profits in each year are obtained by multiplying the survival rate by the total margin per customer net of catalog mailing costs. For line E in Year One, the acquisition cost ($17.5) is subtracted from the sum of expected profits per customer. In subsequent years, the cumulative profits per customer of the previous year are added to the total expected profit per customer of that year.
A.
B.
C.
D.
E.
Frequent Buyers
Year One
Year Two
Margin on each purchase
Survival rate
Cost of mailing catalogs
Total expected profit per customer
Cumulative profits per customer
(net of acquisition costs)
$10
100%
.5*12=$6
2*10-6=$14
$10
75%
$6
.75*(20-6)=$10.5
$(3.5)
$7
3 For frequent buyers: .75(5-1) =.754 ≅.30, and for occasional buyers: .5(5-1) =.54 ≅.06. See also line B of Figure B.
3
89
503-019
A.
B.
C.
D.
E.
Customer Profitability and Lifetime Value
Occasional Buyers
Year One
Year Two
Year Three
Margin on each purchase
Survival rate
Cost of mailing catalogs
Total expected profit per customer
Cumulative profits per customer
(net of acquisition costs)
$16
100%
.5*12=$6
16-6=$10
$16
50%
.5*4=$2
.5*(16-2)=$7
$16
25%
$2
.25*(16-2)=$3.5
$(7.5)
$(0.5)
$3
From the analysis above, the direct retailer finds that frequent buyers become profitable by the end of year two. Occasional buyers become profitable by the end of year three.
Additional Comments:
a) In the above calculations, the discount rate has been ignored. The discount rate allows the seller to control for the fact that money received in the future is not equal to the value of money received today. Money today can be invested and interest on it can be earned. To adjust for discounting, total expected profits for each customer should be deflated for each a year by a factor of D = (1 + i)ª, where i is the interest rate, and a the number of years you wait to receive the money. The discount rate will be taken into account in the next part of the example. b) For simplicity, the analysis in line D assumes the retailer would cease sending catalogs to customers who drop out, in the same year they stop making purchases. In reality, a firm may send out the full number of planned catalogs prior to delisting a customer who does not make purchases in a given year. If this is the case, one should adjust line D and multiply the cost of mailing out catalogs by the survival rate of the preceding year (and not the current year). c)
In some cases, gross margins are not explicitly given and one has to establish them by incorporating all costs associated with a purchase (such as, delivery, payment to manufacturer, and processing fees). In other cases, the “net contribution” appearing on a company’s income statement may need to be adjusted for sales that eventually didn’t take place due to returns and cancellations (by multiplying contribution by (1-% sales returned)).
If net contribution includes costs of promotions and communications that are then separated out in the analysis, such as cost of mailing catalogs in the example above, those need to be excluded from the net contribution percentage.
Part III—Lifetime Value Analysis
Besides knowing how many years it will take to start making profits, it is obviously of interest for the catalog retailer to establish the total expected profit stream arising from each customer over the lifetime of his or her relationship with the company.
In the analysis below, we assume an interest rate of 10% for purposes of discounting future revenue streams. Expected profits are recognized at the end of each year.
4
90
Customer Profitability and Lifetime Value
Figure B
503-019
Lifetime Value Analysis for Frequent and Occasional Buyers
In line F, the cumulative profits per customer are determined in Year One by subtracting the acquisition costs
($17.5) from the net present value of profits. In subsequent years, the net present value of profits is added to the previous year’s cumulative profits per customer.
Frequent Buyers
A.
B.
C.
D.
E.
F.
Margin on each purchase
Survival rate
Cost of mailing catalogs
Total expected profit per customer Net present value of profits per customer (discount rate applied)
Cumulative profits per customer
(net of acquisition costs)
Occasional Buyers
A.
B.
C.
D.
E.
F.
Margin on each purchase
Survival rate
Cost of mailing catalogs
Total expected profit per customer Net present value of profits per customer (discount rate applied)
Cumulative profits per customer
(net of acquisition costs)
Year
One
Two
Three
Four
Five
Six
Seven
Eight
$10
100%
$6
$10
75%
$6
$10
56%
$6
$10
42%
$6
$10
32%
$6
$10
24%
$6
$10
18%
$6
$10
13%
$6
$14
$10.5
$7.9
$5.9
$4.4
$3.3
$2.5
$1.9
$12.7
$8.7
$5.9
$4
$2.8
$1.9
$1.3
$0.9
$(4.8)
$3.9
$9.8
$13.8
$16.6
$18.5
$19.8
$20.7
Year
One
Two
Three
Four
Five
$16
100%
$6
$16
50%
$2
$16
25%
$2
$16
12.5%
$2
$16
6.25%
$2
$10
$7
$3.5
$1.75
$.88
$9.1
$5.8
$2.6
$1.2
$0.5
$(8.4)
$(2.6)
$0
$1.2
$1.7
In computing lifetime value as depicted in Figure B, the retailer has assumed a relationship with the customer until the point where net present value of profits become less than $1 for the first time.4
This extends our previous break-even analysis across multiple years. As can be seen, a frequent buyer has an eight-year time span and generates a little over $20 of profits. An occasional buyer has an expected useful life of five years and generates about $2 in profits. The retailer may decide to send more (or fewer) catalogs or targeted promotions to these buyers to see whether they increase profitability or, alternatively, may decide they are not worthwhile pursuing and drop them once identified. To calculate aggregate profits, the retailer needs to know the size of each segment. This can be determined by the size of the pool from which customers will be acquired and the probability an acquired customer is a frequent/occasional buyer. Note that cumulative profitability of each customer differs from Figure A, given that discounting has been incorporated.
4 Clearly one could use other criteria, such as when the survival rate drops below some percentage. In some cases, the duration
of the relationship is known with greater accuracy in advance (such as reaching a certain age).
5
91
503-019
Customer Profitability and Lifetime Value
One can summarize the analysis presented in Figure B by formally writing an expression for customer lifetime value (CLV) as follows:
(M a − ca )r (a −1)
−
CLV =
(
1 + i )a a =1
N
1
∑
AC
where:
N = the number of years over which the relationship is calculated.
Ma = the margin the customer generates in year a. ca = the cost of marketing communications or promotions targeted to the customer in year a.
(a-1)
r = the retention rate (r
is the survival rate for year a).
i = the interest rate.
AC = the acquisition cost.
If Ma and ca are relatively fixed across periods, one can simplify the above expression by assuming an infinite economic life (i.e., letting N→∞), which leads to:
CLV2=
M −c
− AC
1− r + i
Using this approach, it is easy to establish that the CLV of frequent buyers is
20 − 6
− 17.5 = $22.5
1 − .75 + .1
Unless the retention rate is exceptionally high, this simplification produces results that are very close to the more precise formulation. The more years used in Figure B (or CLV1), the lower the discrepancy. Additional Comments:
a)
It is also possible to use a variation of CLV2 for occasional buyers. There is a need to adjust for year one because the cost of mailing catalogs is $6 instead of $2 for each subsequent year. This
16 − 2
⎛ 16 − 2 16 − 6 ⎞ yields −
− 17.5 − ⎜
⎟ = $ 2 .2
1 − .5 + .1
⎝ 1 + .1 1 + .1 ⎠
b)
If inflationary pressures are expected, one can adjust the expression for CLV1 by multiplying
(a-1)
the numerator by (1+u) , where u is the annual level of inflation. In CLV2, the denominator would become (1-r∗(1+u)+i).
c)
In some business contexts, the relevant time frame between purchases may be more than a single year (for example, with industrial clients upgrading computer hardware every three years). In these settings, calculating CLV may be easier when working with periods that recur every x number of years (x= time between purchases) than with annual values. The analysis would then require transforming the discount rate and inflation rate (that are typically given
6
92
Customer Profitability and Lifetime Value
503-019
in annual values), to span the duration of the period. Thus, in the expression for CLV1, a would be the number of periods and one would replace i by (1+i)x-1.
d)
In some business contexts Ma and/or ca, the per-period profits and costs of serving customers, may not be fixed over the entire lifetime of the customer. For example, as mechanical products age and require more maintenance, the servicing agent can expect to see increased profits from a customer. Annual average profits per customer may not highlight these changes. These make it more difficult to use the simplified expression CLV2. The same is true for the retention rate r, which may increase the more purchases an individual makes with the company. By contrast, competitor poaching and harsher economic conditions usually negatively impact the retention likelihood.
Strategic Implications of Customer Lifetime Value Analysis
The benefit from CLV calculations is two-fold: understanding the potential value of customers and prompting firms to learn more about the patterns of individuals or groups of customers. This information allows the firm to devise optimal strategies for each customer, eliminate wasteful costs, and create a long-term perspective of the potential relationship with customers. Firms can tailor strategies to deal with different customer segments that exhibit differences in buying characteristics at any given time, and they can also customize different strategies for the same customer depending on the stage of relationship between the customer and the firm.
Measuring CLV typically involves the use of historical sales data to allow detailed analysis of the profits customers are likely to produce across their “lifetime” (the time a company is likely to retain a customer). When a business is familiar with the historical buying patterns of its customers—and future customers are assumed to produce predictable levels of loyalty—then CLV can be used to estimate the likely profitability of those customers. As stated above, when a company has calculated
CLV, it is in a position to evaluate its customer base and devise marketing strategies for individuals or groups of customers. Common actions include:
a) “Firing” Customers
A company may find that only a fraction of its customers are profitable. Aside from attempting to make marginal customers become profitable, the company may be motivated to lose a percentage of the least-performing individuals. In some cases, this can be achieved by simply raising prices for the less profitable products. Best customers typically outspend others considerably, with a ratio of 15 to 1 in some industries. There is a competitive advantage for firms that can retain high value customers with minimal costs, and detailed information about customer profitability allows a firm to identify weaker customers. For purposes of firing customers, firms may also attempt to identify customers who create substantial costs, such as those making regular product returns or demanding increased levels of service.
b) Rewarding Customers
Likewise, a firm may choose to reward customers with discount vouchers or preferential services.
Identifying a firm’s best customers and investing in them can lead to higher loyalty (thus increasing their retention rate) and to improve sales.
7
93
503-019
Customer Profitability and Lifetime Value
c) Identifying Cross-Selling Opportunities
With detailed information about the interests and shopping patterns of individual customers, firms can identify opportunities to offer additional or related products (either separately or in the form of package deals). In some cases, firms may choose to proactively leverage collaborators whose products and competencies complement their own.
CLV and Relationship Marketing
In closing, it is worth emphasizing that the relationship between the customer and the firm typically evolves over time and is not static. To enhance the mutual value created by the relationship, it may be important for the firm to recognize the different phases of the customer life cycle. One subcategorization that can be useful for lifetime value analysis, suggested by Blattberg, Getz and Thomas in Customer Equity: building and managing relationships as valuable assets. (Boston, MA: Harvard
Business School Press, 2001), identifies five life cycle stages:
Stage 1: Prospects
Marketing effort is directed at attracting potential customers, who at this stage primarily develop expectations about the firm and its competencies. This stage is critical, therefore, for determining the long-term satisfaction and retention of a customer. If the firm creates expectations that are exceedingly high, customers may be easily acquired but will not be satisfied or retained, affecting both per-period revenues and survival rates.
Stage 2: First-Time Buyers
After making their first purchase, customers have some experience to assess the value created by the firm’s products and services for them—the risk of defection is typically high during this stage. For industries with short purchase cycles, satisfaction during this stage is critical for establishing repeat buyers. Stage 3: Early Repeat Buyers
Customers who make a repeat purchase are more likely to buy again than first-time buyers. As their confidence in the firm grows, their perception of value from the product increases and the rate of defection is reduced.
Stage 4: Core Customers
The rate of defection in this stage is typically low, and customers will not defect unless they encounter a high level of dissatisfaction.
Stage 5: Core Defectors
Many reasons, including the arrival of competitors with new products or enticing promotions and inconsistent delivery of service/quality by the firm, eventually cause some customers to switch suppliers. In some situations, external factors do not allow a firm to react to the loss of the customer,
8
94
Customer Profitability and Lifetime Value
503-019
while in other situations a firm can bring back a defector if the problem can be identified and rectified. Relationship marketing assumes the possibility of creating an ongoing relationship with a customer, and that customer satisfaction will be strong enough to create loyal or repeat buyers. In these situations, firms strive to find ways to increase customer satisfaction: provision of quality, additional services, and rewards for loyalty and repeat purchases. Satisfied customers tend to be less price sensitive and are likely to remain loyal for a longer period.
Evolving customer needs, competitors’ actions, opportunities to leverage collaborators, and changing economic and social conditions can all potentially affect the relationship between the firm and its customers. Customer lifetime value analysis, and the constant re-evaluation of the parameters determining it, presents the company with a framework to optimally respond to these changes.
References
Berger, P. and Nasr, N. “Customer Lifetime Value: Marketing Models and Applications.” Journal of
Interactive Marketing. Volume 12, No. 1 (1998).
Blattberg, R., Getz, G. and Thomas, J. Customer Equity: building and managing relationships as valuable assets. Boston, MA: Harvard Business School Press, 2001.
Blattberg, R. and Deighton, J. “Manage Marketing by the Customer Equity Test.” Harvard Business
Review (July-August, 1996).
Dwyer, R. “Customer Lifetime Valuation to Support Marketing Decision Making.” Journal of Direct
Marketing. Volume 11, No. 4 (1997). th Kotler, P. Marketing Management: analysis, planning, implementation, and control. 9 Edition. Upper
Saddle River, NJ: Prentice Hall, 1997. th Kotler, P. and Armstrong, G. Principles of Marketing. 9 Edition. Upper Saddle River, NJ: Prentice Hall,
2001.
Rust, R., Zeithaml, V., and Lemon, K. Driving customer equity: how customer lifetime value is reshaping corporate strategy. New York: Free Press, 2000.
Schnaars, S. Marketing Strategy: A Customer-Driven Approach. New York: Free Press, 1991.
9
95
96
9-500-024
REV: JANUARY 15, 2002
JOHN DEIGHTON
The Brita Products Company
In 1987 when Charlie Couric saw his first Brita pitcher he thought, “A homemade alternative to bottled water!” Here was a product that, with the right marketing support, could be very successful.
Couric, a marketing executive with the Clorox Company charged with finding new business ideas, had been browsing health food stores in California when he came across the quirky home water pitcher-and-filter system made by a small German company, Brita GmbH. He proposed that Clorox acquire the right to market Brita in the USA, and in 1988 they did so. Couric reflected:
In the early years we had to beg the corporation to invest. Some of my colleagues viewed the pitcher as another waffle iron - used once and then tossed into the basement. We saw it differently. We looked at the repeat purchasing of filters, and to us the strategy was obvious.
This was a race to put a pitcher on every kitchen countertop, at a loss if necessary.
Clorox supported Couric’s deficit-spending proposal, and a decade later Brita had grown to become one of Clorox’s biggest brands. It had rewarded Couric’s faith, spearheading the growth of a home water filtration industry in the United States. More than 17 million Brita pitchers had been sold, and each pitcher sale started a stream of filter sales. The Brita brand was generating close to
$200 million revenues a year.
Now, in 1999, Couric was keeping an eye on a different water purification product launched by a small competitor, PUR. It was a filter that screwed onto kitchen faucets.1 Clorox had its own version of the faucet-mounted filter ready for launch, and again a debate had developed over whether to deficit-spend. Some counseled that the faucet-mount had the power to disrupt the pitcher product, and Brita had no choice but to pour money into another race to build another installed base, this time in faucet-mounted products. Others argued that faucet-mounts served a different niche of the water purification market from pitchers, and the two could live side-by side. A third group argued that
Brita should do nothing to foster faucet-mounts. Its priority, they argued, was to invest to defend its installed base of pitchers and the associated filter revenue stream. Money spent on promoting a faucet-mount would only erode the pitcher base and interrupt its stream of filter revenues. They pointed out that PUR was a small, loss-making firm, too weak to succeed at creating a new category, particularly when the early adopters of home water filtration all had pitchers in their homes.
1 Faucet-mounted filters were not themselves a new product—Teledyne had sold one since the 1960s without much success.
However, the introduction of solid carbon block technology in 1995 improved performance and increased consumer interest.
________________________________________________________________________________________________________________
Professor John Deighton prepared this case. HBS cases are developed solely as the basis for class discussion. Cases are not intended to serve as endorsements, sources of primary data, or illustrations of effective or ineffective management.
Copyright © 1999 President and Fellows of Harvard College. To order copies or request permission to reproduce materials, call 1-800-545-7685, write Harvard Business School Publishing, Boston, MA 02163, or go to http://www.hbsp.harvard.edu. No part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying, recording, or otherwise—without the permission of Harvard Business School.
97
500-024
The Brita Products Company
The Clorox Company
The Clorox Company was a major manufacturer and marketer of laundry additives, household cleaners, charcoal, auto care products, cat litter and home water purifiers. In January 1999, Clorox bought First Brands, a $1.2 billion manufacturer of plastic wraps and bags, auto care products, cat litter and home fireplace products. Revenues of the two companies combined would have been $3.9 billion in 1998. Some of the well-known U.S. consumer brands that would come under common ownership following the merger were:
Clorox
First Brands
Armor All car care products
STP automotive products
Fresh Step cat litter
Scoop Away cat litter
S.O.S. steel wool pads
Ever Clean cat litter
Hidden Valley salad dressing
Johnny Cat cat litter
Kingsford charcoal
StarterLogg fire starters
Clorox laundry bleach
HearthLogg fire logs
Soft Scrub cleaners
Glad plastic wraps and bags
Brita water filtration systems
Formula 409 spray cleaner
Tilex cleaners
Pine-Sol cleaner
Liquid Plumr
Under Chairman and CEO G. Craig Sullivan, Clorox followed a strategy of building dominant brands, pursuing international expansion and acquiring promising businesses. Some 85% of Clorox brands were first or second in their categories. Sales beyond the United States would reach 20% of revenues after the First Brands merger, up from 18% for Clorox alone.
The Brita Products Company
Brita GmbH, a family-owned corporation headquartered in Tanusstein, Germany, made a variety of industrial and consumer water filtration products. Before Couric called, it had struggled without much success to sell its home water filtration system in the United States, most recently through a
Canadian agent. After vigorous negotiation, in September 1988 it agreed to let Clorox form a subsidiary, Brita USA, to be the sole U.S. distributor of Brita products. Clorox would buy filters from
Brita GmbH and design and make its own pitchers. Couric became President and General Manager of Brita USA.
For four years, Brita USA struggled. The costs of building distribution, designing products, and promoting the concept dwarfed the small base of sales. Couric persevered, however, because early surveys of users suggested to him that a Brita customer would have a remarkable lifetime value.
Each pitcher sale would start a flow of filter sales. Over 80% of pitcher buyers were still using the product a year later and many were telling friends to try it and were giving it as gifts. In the 1990s, the product took hold like crabgrass. By 1999, an estimated 13% to 15% of the 103 million households in the United States were using a Brita pitcher. Brita had created a home water purification industry worth $350 million at retail, and held a 70% revenue share.
2
98
The Brita Products Company
500-024
The Product
The Brita pour-through filtration system comprised a two-compartment pitcher and a replaceable filter (Exhibit 1). Tap water, poured into the upper compartment, flowed under gravity through the filter into the lower compartment, filling it in about five minutes.
The filter had two elements. Activated carbon reduced chlorine, sediment and odors, and an ionexchange resin removed any heavy metals such as lead, copper, mercury and cadmium, as well as temporary water hardness (calcium and magnesium). The benefits were threefold. Filtered water tasted better, it did not deposit scale when boiled, and, to the extent that it might have contained harmful heavy metals, they were extracted. The filter did not screen out microorganisms such as cryptosporidium and giardia, two sources of gastro-intestinal illness that were potentially fatal to people with compromised immune systems.
The pitcher system was sold with a single filter in place. Filters required replacement every two months or after filtering 40 gallons of water. Brita supplied calendar stickers to help users track when a filter needed replacement. Filters were sold in packs of one, three and five.
Consumer Attitudes and Behavior
Over the decade of the 1990s, the safety of tap water became a topic of growing concern to U.S. households. In one well-publicized case, two wells supplying the drinking water for the Boston suburb of East Woburn, MA were found to be contaminated with industrial solvents, coincident with a number of cases of leukemia. The incident was the subject of a book and a 1998 motion picture, ‘A
Civil Action.’ In that year the U.S. Environmental Protection Agency declared that about 10% of the sediment under U.S. surface waters is "sufficiently contaminated" with toxic pollutants to pose a health threat to humans and wildlife. Later that year, Congress began requiring municipal water authorities to say when contaminant levels exceeded federal regulations. In Milwaukee in 1993,
403,000 people were made sick and 111 died when the parasite cryptosporidium entered the municipal water supply. By the end of the decade, a poll by USA TODAY, CNN and Gallup found
2
that 47% of respondents preferred not to drink water straight from the tap.
Sales of bottled water from U.S. supermarkets and home delivery services grew rapidly during the decade. By 1997, bottled water made up 8% of all the liquids that people paid to drink and was the industry’s fastest growing category (Exhibit 2). In that year, 45% of households bought still (noncarbonated) water in supermarkets and 27% bought carbonated water. The average price paid for still water was about a dollar per 128-ounce container, and carbonated brands averaged about three dollars (Exhibit 3).
3
A 1998 survey found that two-thirds of Americans claimed to be familiar with the expert recommendation to drink eight, eight-ounce servings of water a day, yet only one in five drank that quantity of water and 44 percent drank three or fewer water servings daily.
2 Eisler, Peter, Barbara Hansen and Aaron Davis. “Lax oversight raises tap water risks.” USA TODAY, October 21, 1998, p.
15A.
3 Yankelovich Partners conducted a survey of 3,003 Americans for the Nutrition Information Center at The New York Hospital
- Cornell Medical Center and the International Bottled Water Association. The study is described at http://www.bottledwaterweb.com/news/news3.html. 3
99
500-024
The Brita Products Company
A 1999 survey4 found that 72% of all respondents, and 89% of young adults, voiced some concern about the quality of their household’s water supply. A majority of households used either bottled water or some water purification system to limit their exposure to public water supplies. The number taking no precautions declined from 47% in 1995 to 35% in 1999 (Exhibit 4).
Market Performance
Brita used the term ‘systems’ to refer to pitchers and faucet-mounted units, and ‘filters’ for the replacement filters. System sales were sluggish for the first four years after launch, but filter sales grew more rapidly (Exhibit 5). In the early years Couric compared performance in the United States to the first years of the product’s life under other Brita distributors in Canada and the United
Kingdom, and the similarity of the profiles gave him, as early as 1991, the confidence to persevere:
We saw that trial in the United States in the early 1990s was running between the Canadian and the U.K. levels. Close to 25% of buyers told us that they had given a Brita pitcher as a gift.
Another reassuring sign was that surveys were finding that more than 80% of those who had tried the pitcher were still using it a year later. The same surveys reported that they were buying 2 or 3 filters a year. Each year we tried to relate filter sales to past pitcher sales. We found that when we estimated our installed base at 80% of those who had bought a system in the previous five years, and assumed that the installed base bought 2.5 filters a year, the resulting forecasts of filter sales each year were close to reality.
Management tracked system market share in units and filter share in dollars (Exhibit 6). Brita’s share of combined system and filter market revenues had been steady in the range of 65% to 75% from 1995 to 1999. System unit shares were far more volatile. In July 1998, for example, system sales doubled over the previous year and Brita’s share increased ten points in response to a so-called
‘bogo’ (buy one, get one free) promotion on pitchers, intended to pre-empt a PUR competitive launch.
Filter sales had been less responsive to sales promotion activity by manufacturers.
Brita bought pitchers from contract manufacturers at a cost per unit of $7.80. Filters were purchased from a manufacturing plant owned jointly by Clorox and Brita GmbH for $2.05 per unit, inclusive of a 3% royalty to Brita GmbH. (Profits from this plant were not material.) At the prices
Brita charged retailers in 1999, pitchers earned a contribution to fixed costs of 48.6% of Brita’s net revenue, and filters earned 50.0%. After advertising and trade promotions, Brita USA earned a net return on sales of 24%, the highest of all Clorox business units. Although advertising spending worked to the benefit of both pitchers and filters, trade promotions were used mainly to secure trade support for pitchers. Exhibit 7 summarizes the income statement of the brand in 1998.
Distribution
At inception, Brita’s main retail outlet had been a health foods chain. Its competitors were in stores that stocked housewares, like Sears and Walmart. Couric believed, however, that the product would flourish in Clorox’s traditional base, grocery and drug outlets, and drove distribution in that direction. 4 “1999 National Consumer Water Quality Report,” Lisle, IL: Water Quality Association, 1999.
4
100
The Brita Products Company
500-024
Brita distribution
1992
1998
Department stores
27%
13%
Mass merchandisers
31%
34%
Grocery stores
11%
14%
Club stores
31%
21%
Drug stores
-
12%
This pattern of channel evolution, in which high margin retailers like department stores pioneered a new category, only to lose share to lower margin retailers, was known among marketers generally as a “class to mass” strategy. Couric explained his strategy:
Our version of “class to mass” had three elements. We wanted to be established in class, first in mass, and alone in grocery. So we created a line of upscale pitchers called Ultra for department stores, appropriate for their 35% mark-up structure. We sold the standard pitcher, inherited from Germany but manufactured locally, in mass merchants like Target and Walmart and in drug and grocery stores. They marked it up 25%. We designed a bonus pack system and a 5-pack of filters to appeal to club retailers when they became important in the early
1990s.
Keeping these various classes of trade happy is an enormous challenge. We are continuously building the system and seeing cracks appear. Eventually perhaps we’ll be driven out of ‘class,’ but we aim to make it last as long as possible. We need the breadth of distribution and variety of products to support our $30 million advertising budget, and to provide a channel for introducing and establishing future new products.
One way we attempt to keep the peace among our classes of trade is by insisting that no retailer advertise a Brita line at below the price that we set. We call it a MAP
(minimum advertised price) policy. We reimburse retailers for featuring our products in their display advertising, but if they feature us at below the MAP, then we won’t pay. The only exception we make is on the standard pitcher. We let them deal as much as they like on that item.
Positioning and Advertising
Brita’s advertising in the United States emphasized a taste benefit. Couric explained:
Initially people had no clue about the concept of a pitcher product. I remember on the cab ride back to the airport after our first trade show in Chicago I explained the concept in lengthy detail to the cab driver and when I was finished he said, ‘You screw it onto the faucet, right?” I realized then that we had to tell people how the product worked, so we split the advertising message 50/50 between how it works and how it tastes. Today, now that our product is more well known, we are able to be more focused about the taste benefit.
We decided on taste for three reasons. First, research showed that when we talk taste, we get a health benefits halo. When we talk health, we don’t communicate the idea of taste.
Second, we noted that the whole bottled water industry had been built without reference to health. Third, I wanted to be at the top of the mountain. I didn't want competitors overtaking us. If we focussed on lead removal, say at 93%, someone else could claim to take out 95%.
5
101
500-024
The Brita Products Company
With taste, we could say it first, say it loudest, and we could own the benefit. By now, with
$100 million of cumulative advertising on the taste claim behind us, we are impossible to dislodge. When we started eleven years ago, the water filtration category had low credibility. It was being investigated all the time for improper or false claims. We didn’t want to get into a claims war. Our advertising needed to be pure and simple. We showed mountain streams, waterfalls and the outdoors. We promised clear, crisp, refreshing water, which is what we delivered.
Today we own the waterfall imagery.
Competition
Brita’s success attracted competitors in droves. Among the brand names that entered the market were Culligan, Electrolux, Sunbeam, Kenwood, Corning, Melitta, PUR, Rubbermaid, Teledyne, Omni and Mr. Coffee. None had succeeded in challenging Brita’s leadership, which remained in the range of 65% to 75% across the decade. In 1999, the only competitor with double digit market share was
PUR, the brand of a small, publicly held U.S. corporation, Recovery Engineering. This company made, in addition to water filters, a line of portable drinking water systems for outdoor enthusiasts and desalinators for marine and military use.
At the January 1998 International Housewares Show in Chicago, a dozen manufacturers unveiled new water filtration products or extensions to existing lines. The products appeared to be designed to attack niches not currently served by Brita. "Number 1 is always going to be under attack," Brian
Barton, brand manager at Rubbermaid, told the press covering the trade show. "When you have 80%
5
of the market share there's only one way to go and that's south."
Several products took health and safety positions in reaction to Brita's taste appeal. Number 2 brand, PUR, announced that it would spend $40 million in advertising and promotion to support its line of faucet-mount and pitcher filters. The spend included a $15 million outlay for PUR Plus, a new pitcher system touted as the most technologically-advanced to date. The PUR filter would remove contaminants such as cryptosporidium and giardia. PUR representatives described a promotional program that would begin with a six month infomercial running on national cable television, to be followed by a schedule of 30-second spots on cable and spot network TV that would point out the differences between PUR and current pitchers. Sunbeam, known for blenders and toasters, announced the launch of Fresh Source, a product that removed microbiological cysts as well as chlorine and lead. Sunbeam would back the product with an estimated $10 million of advertising.
Number 3 manufacturer Teledyne, unveiled a faucet-mount product at the show.
At the trade show, Brian Sullivan, president and CEO of Recovery Engineering, Minneapolis was quoted as saying, "If you ask consumers if they want more contaminants taken out or less, they'll say more. People will pay more for a higher-performing product." PUR’s pitcher was considerably more expensive than Brita’s standard product. A Brita representative responded, "The way we see the market, this business is geared more towards taste. Consumers are interested in taste. Bacteria is way
6
down the list."
In a $30 million ad campaign unveiled in the month of the Chicago show, Brita did not mention harmful impurities. A TV spot dubbed "There Was a Time" features shots of rushing streams set
5 Mehegan, Sean. “Sunbeam, recovery loading up $$ to take on Brita in water filtration.” Brandweek, Vol. 39, Iss. 3, January 19,
1998 p. 12.
6 Op cit., p. 12.
6
102
The Brita Products Company
500-024
against a backdrop of mountains and a dark, brooding sky. "There was a time when it was perfect," the voiceover says. "You can have this taste . . . again."
Rubbermaid had launched a low priced product in 1997, which used a technology similar to
Brita’s while attacking it on price/performance. Rubbermaid claimed that its filter could cleanse 800
8-oz. glasses of water versus 560 for Brita at the same price. In 1998 the company repackaged its pitcher product and announced a portable 16-oz. bottle with a carbon filter built into the cap.
Rubbermaid had not advertised its pitcher in 1997, and sales had been disappointing. At the trade show they pledged to step up promotion under a new team led by Cathryn Rings, the former head of
Procter & Gamble's Max Factor cosmetics business. Rings announced at the trade show, "We're going to try some classic P&G marketing.”
The Faucet Mounted Filter Entry
Prior to 1995, Brita executives expressed little interest in faucet mounted water filtration systems.
Of the 50 countries in which Brita GmbH did business, only Japan had a significant faucet business, and that was attributed to space constraints in Japanese kitchens. In 1994 and 1995, however, they saw a faucet segment forming. Recovery Engineering Inc. launched a faucet mounted product under the PUR brand name with some success. In 1995, Brita hired an outside design company to design a faucet mount.
Functionally, water from a pitcher was different from water filtered through a faucet-mounted filter. In favor of pitchers, they were usually stored in refrigerators, so pitcher water was cold while faucet water was not. In those parts of the country where tap water was ‘hard’ and left scale deposits and scum when boiled, only pitcher filters but not faucet filters would eliminate hardness. In favor of faucet mounts, the water passed through at higher pressure than through pitcher filters, so finer filters could be used that could screen for microorganisms and offer protection against cryptosporidium and giardia. Also, pitcher-filtered water tasted crisper, with lower pH. Finally, faucet-filtered water cost significantly less per glass because the filter lasted longer. Where pitcher water cost 15 to 20 cents per gallon, faucet-filtered water cost perhaps half.
Were these differences significant to consumers? Did Brita stand for good tasting water, or how you get it? The Brita team debated whether the faucet mount would be perceived as another way to deliver Brita water. Or would the consumer decide that they were buying something quite different, perhaps even so different that some might consider it a good idea to own both a pitcher and a faucet mount? As Brita’s filtration technology played no part in the faucet mount design, Clorox was not obliged to use the Brita name on this product. If it did so, however, it was required to pay Brita GmbH a royalty that, under the 1988 agreement, would be between 3% and 4% of sales depending on the magnitude of sales. It would also be bound by the non-compete clause of the comprehensive agreement that limited sales of products with the Brita name to North America. Conversations with the retail trade, however, revealed a distinct preference for carrying the faucet mount under a wellknown name like Brita.
The direct cost of the faucet mount system was estimated to be $15.00 and the direct cost of a replacement faucet filter would be $3.00. Pitcher filters could not be used in faucet mounts.
Would a faucet mount cannibalize pitcher and pitcher filter sales? Perhaps, some speculated, the pitcher was a starter product, and customers who had learned to go back to drinking tap water would graduate up to the more convenient and sophisticated faucet unit. To explore these and other
7
103
500-024
The Brita Products Company
7
questions, Clorox commissioned a simulated test market from ACNielsen Vantis, a division of the
ACNielsen BASES group.
Market Simulation Study for the Faucet Filter
8
In the ACNielsen Vantis study, 567 respondents, characterized as water-involved and drawn from eight markets across the United States, were intercepted in shopping malls, brought to rooms containing simulated retail store shelves, and asked to choose from a display of water filtration systems. Respondents were assigned to one of three rooms. Each room displayed ten products currently available in the market. In addition, two displayed a prototype of the Brita Faucet Filter System, one priced at $34.99 and the other at $39.99. The third had no Brita Faucet Filter System on display, to serve as the control cell of the experiment. In the first two rooms, subjects saw print advertising for
Brita and PUR faucet mounted filters.
Consumers were first asked to rate how likely it was that they would buy any of the displayed items in the next two months. They were then asked to identify their first, second and third choice of item. Finally, they were asked a series of questions about the test product, the Brita Faucet Filter
System. Vantis offered the Brita team the following conclusions from the study:
•
The Faucet Filter increased the likelihood of buying a product from the Brita line.
•
However it did not increase interest in the filtration category as a whole, so that the combined pitcher and faucet-mounted market was not expected to expand.
•
Though higher priced, the Brita Faucet Filter generated similar levels of purchase intention to the Brita Spacesaver Pitcher.
•
About half the Brita pitcher owners who bought the Faucet Filter system would continue to use the pitcher in conjunction with the faucet product.
•
Both Brita and PUR’s faucet filters were perceived to be superior to the Brita pitcher in removing contaminants, and in convenience. However, only Brita’s faucet filter was perceived to improve the water’s taste.
•
Unit sales and perceptions of value for the faucet mount were strong at both the $39.99 and
$34.99 prices, and sales would not be significantly impacted if the PUR price was dropped by
$5.00.
7 Simulated test market studies have a long history, dating back to the 1970s when Alvin J. Silk and Glen Urban began research
at MIT’s Sloan School to seek ways to forecast demand for new products without incurring the costs and public exposure of full-scale test markets. Today simulated test markets are regularly relied on to forecast in-market performance without the need to build production capacity, expose marketing plans to competitive scrutiny, and wait six or twelve months to read results. Their evolution is described in Kevin J. Clancy, Robert S. Shulman and Marianne Wolf, Simulated Test Marketing:
Technology for Launching Successful New Products, New York, NY: Lexington Books, 1994. ACNielsen Vantis serves services and durable goods industries with a brand of simulated test market methodology, BASES, that derives from work that began in the
Pillsbury Company in the 1960s, found a temporary home in Booz-Allen & Hamilton (the name is an acronym for Booz-Allen
Sales Estimation System), was spun off in a leveraged buyout in 1977, and is now a division of the A. C. Nielsen Company.
8 A water-involved respondent was one who either owned a filtration device or bought bottled water, and described him- or
herself as not satisfied with the quality of their water.
8
104
The Brita Products Company
500-024
The study projected that unit sales of the faucet mount in its first year would lie between 350,000 and 1,395,000 units. Half of this volume would come from consumers who would otherwise have bought a Brita pitcher. Whether sales would be at the high or low end of the range depended on how aggressive was Brita’s marketing investment, and how competitors responded. Ten scenarios were generated by combining the following factors and levels:
Low
Consumer advertising
Consumer promotion
Feature price reductions
Other trade spending (displays, racks, etc.)
$5.4 million
$2.0 million
$1.8 million
$3.2 million
High
$11.1 million
$3.0 million
$2.3 million
$6.1 million
Very High
$15 million
$4 million
$2.8 million
$9 million
The ten scenarios gave rise to the following ten sales forecasts:
Scenario:
1
2
3
4
5
6
7
8
9
10
Minimum
Advertised Price
($5.00 off List)
$34.99
$34.99
$29.99
$34.99
$29.99
$29.99
$34.99
$34.99
$29.99
$29.99
Consumer
Promotion and
Trade Spending
Consumer
Advertising
Competitive
Pricing
Low
Low
Low
High
High
High
High
Very High
Very High
Very High
Low
Low
Low
High
High
High
Very High
Very High
High
Very High
Low
Current
Low
Low
Low
Current
Current
Current
Current
Low
First Year Unit
Sales Forecast
340,000
350,000
395,000
970,000
1,125,000
1,160,000
1,205,000
1,245,000
1,350,000
1,395,000
Each scenario resulted in a sales and income forecast. For example, scenario 2 led to the following forecast: Total households
Product awareness resulting from $5.4 million advertising
Distribution (% of market reached)
List price (30% of sales assumed to occur at list)
Feature price (70% of sales assumed to occur at the feature price)
Trade promotion
Consumer promotion
Competitive pricing
Total sales (units)
75.86 million
13%
72%
$39.99
$34.99
$3.2 million (low)
$2.0 million (low) current levels
350,000
Brita USA had not asked Vantis for a forecast of sales of replacement filters. The proposed product had an LED filter replacement indicator, which would likely increase compliance with filter replacement recommendations. Each filter would treat 100 gallons of water, about four months’ output from a typical kitchen faucet, before the indicator would signal that it was due for replacement. 9
105
500-024
The Brita Products Company
Couric’s Decision
Couric prepared to call his marketing team together to hear their views on how to take the Brita brand forward. He anticipated that he would hear three points of view: keep the focus on building the installed base of pitchers, shift the budget to encourage the installed base to buy more filters, or put the weight of resources behind building a whole new installed base in faucet-mounts.
He saw many demands on the Brita marketing budget besides the faucet-mount product launch.
Household pitcher penetration was slowing, and yet six out of seven households did not have one.
Could there be segments who had not responded to the broad appeals of the first decade, but who might well respond to more targeted communication efforts—specific appeals to singles and to parents of young children, for example? Perhaps investing in direct mail or other highly targeted marketing tools could cultivate demand in these niches. Then there was the filter opportunity. Brita had never invested in the direct cultivation of filter demand, beyond in-store promotion.
On his desk in the corner office were Recovery Engineering’s published financial results for the quarter ending January 3, 1999. Its quarterly sales were up $1 million on the previous quarter to
$19.5 million, but its net loss had more than doubled to $7.2 million. Its stock was trading at $10, down from $35 in mid 1998. Recovery Engineering had raised capital in an Initial Public Offering in
1997 on the claim that it had a technological edge over Brita. To be sure, PUR had been first to market with a number of new features: the first cryptosporum filter for pitchers, the first mechanical device to indicate when to replace a filter, and the first widely distributed faucet filter. But as Couric weighed how much urgency to put behind the faucet-mount launch, he found comfort in PUR’s flow of red ink.
A Fax and a Phone Call
Couric’s fax began transmitting. Over the line came a report from the Clorox field sales director, with a sketch of a display that had been seen that morning in the Schaumburg, IL branch of Target
Stores, in the store’s large water filtration section. A sign over the display had read:
Which water filtration product is right for me?
How do different water filters give me great tasting water and protect my family?
Choose your level of protection:
•
••
•••
••••
Lead, chlorine.
Lead, chlorine, cryptosporidium, guardia.
Lead, chlorine, cryptosporidium, guardia, Lindane (a pesticide), Atrazine (a herbicide), asbestos.
Lead, chlorine, cryptosporidium, guardia, Lindane (a pesticide), Atrazine (a herbicide), asbestos, benzene, TTHMC.
All product claims are NSF® certified to national public health standards.
Beneath the sign he had seen five PUR systems and four Brita systems mounted on identical backing cards labeled with one, two, three or four bullets. No PUR system had fewer than two bullets and the PUR Ultimate Faucet Mount had the maximum, four. Not a single Brita system had more than one bullet.
Simultaneously, Couric’s phone rang. His investment bankers were advising that Procter and
Gamble, the world’s largest consumer products company and Clorox’s most respected competitor, was about to close a deal at $35 per share for control of PUR.
10
106
The Brita Products Company
Exhibit 1
500-024
1993 Advertising
11
107
108
Exhibit 1
(continued)
Brita offers a money-back guarantee (details in box). Brita is available at many fine retailers. For the one nearest you, call 1-800-44-BRITA, or visit our website at www.brita.com. Substances removed may not be in all users’ water. ©1997 The Brita Products Company.
The ingenious filter inside the Brita® Water Filtration Pitcher is one of a kind.
It virtually eliminates lead and chlorine. Dramatically reduces copper, sediment and water hardness. Best of all, it turns tap water into clear, fresh, wonderful water. There’s nothing like the great taste of Brita water. Enjoy.
500-024
-12-
The Brita Products Company
Exhibit 2
500-024
Growth of Segments of the U.S. Beverage Market
Cumulative Growth (Indexed on 1988)
Bottled Water
160%
Fruit Juice
130%
Beer
Soft Drinks
Total Market
Tea
100%
Milk
Coffee
Wine
Spirits
70%
1988
1990
1992
1994
1996
Bottled
Water
Fruit
Juice
Soft
Drinks
Beer
Wine
Per Capita
Consumption
(gallons, 1997)
12.7
15.0
54.6
29.1
1.9
1.2
Segment Size
(Billions of gallons, 1997)
3.4
4.1
14.7
7.9
0.5
0.3
Spirits
Coffee
Tea
Milk
22.6
7.6
20.0
6.1
2.1
5.4
Source: 1999 Beverage Marketing Directory, Mingo Junction, Ohio: Beverage Marketing Corp., 1999.
Total Per Capita Consumption = 144.7 gallons in 1997.
Total Beverage Consumption = 44,500,000,000 gallons in 1997.
13
109
500-024
The Brita Products Company
Exhibit 3
Major Brands in the Bottled Water Category (supermarkets only)
Percent of
Households
Buying
Price per 128 oz. unit
Still Water Brands
Dannon
Arrowhead
Poland Spring
Sparkletts
Chrystal Geyser
Evian
Hinckley & Schmitt
Private label
44.77%
6.27%
4.57%
4.47%
4.05%
3.85%
2.61%
2.38%
17.73%
$1.03
$2.09
$0.95
$1.39
$0.86
$2.13
$5.49
$1.36
$0.68
Carbonated Water Brands
Canada Dry
Schweppes
Vintage
Clearly Canadian
Perrier
Private Label
26.83%
5.69%
4.66%
2.65%
1.75%
1.31%
12.52%
$3.70
$4.97
$5.85
$2.44
$11.66
$9.47
$2.78
Source:
Information Resources Inc. “Marketing Fact Book” January-December 1997,
http://fic.wharton.upenn.edu/iri/factbook .
14
110
The Brita Products Company
Exhibit 4
500-024
Water Quality: Consumer Attitudes and Behavior
From a survey of 1,007 adults conducted between January 14 and 17, 1999. Sample is projectable to all U.S. adults over age 18.
Expressed concerns about household water quality (% of respondents):
Expressed any concern
Health contaminants (net)
Bacteria
Aesthetics (net)
Smell and/or taste
Appearance
Hardness
Sediments
1995
1997
1999
75
54
75
50
61
45
33
38
43
60
42
26
39
40
72
48
9
56
40
26
36
38
Expressed concerns about household water quality by age in 1999 (% of respondents):
Expressed any concerns
Health contaminants
18-24
25-34
35-44
45-54
55-64
89
67
79
56
79
58
70
45
59
36
65+
56
25
Use of water treatment device (% of respondents):
No device used
Bottled water
System
Table-top pitcher
System on faucet
Whole house system
Softener
1995
1997
1999
47
36
27
5
9
10
10
41
37
32
12
11
7
9
35
38
38
16
11
8
9
Use of water treatment device by region of country in 1999 (% of respondents):
North East
Bottled water
Table-top pitcher
System on faucet
Whole house system
Softener
Source:
33
24
9
10
5
North Central
32
12
10
8
21
South
West
42
17
10
7
4
43
12
14
10
6
Water Quality Association: 1999 National Consumer Water Quality Report.
15
111
500-024
The Brita Products Company
Exhibit 5
Brita Unit Sales, 1989 to 1998
Brita Unit Sales (’00)
Systems
Filters
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
171
402
194
581
202
876
302
1292
546
2,205
1,056
4,458
2,030
8,164
3,363
15,246
4,565
23,293
5,266
27,413
Sales of Brita Pitcher Systems and Filters
30000000
25000000
Unit Sales
20000000
Systems
Filters
15000000
10000000
5000000
0
1988
Source:
1989
1990
1991
1992
1993
1994
Company records (approximate and unaudited).
16
112
1995
1996
1997
1998
1999
The Brita Products Company
Exhibit 6
500-024
Retail Market Shares (United States, all retail outlets)
Systems
1992
Pitchers (thousands of units)
Brita
PUR
Rubbermaid
All others
Faucet Mounts (thousands of units)
PUR
Teledyne
375
82%
18%
1,186
23%
1993
640
82%
18%
782
23%
1994
1995
1996
1997
1998
1,405
75%
25%
602
30%
2,636
77%
23%
659
9%
43%
4,381
77%
23%
898
30%
43%
5,689
80%
4%
7%
11%
1,249
67%
27%
6,307
83%
8%
4%
5%
1,291
74%
23%
Filters
Filter sales ($millions, retail)
Brita
Teledyne
PUR
Omni
Sears
Pollonex
Source:
1992
1993
1994
1995
1996
1997
1998
$20.5
32%
25%
0%
12%
7%
7%
$26.5
43%
20%
0%
13%
6%
4%
$38.7
59%
15%
0%
8%
3%
2%
$63.3
65%
10%
1%
8%
2%
1%
$82.3
75%
9%
2%
5%
2%
1%
$116.3
75%
7%
8%
3%
1%
0%
$154.7
75%
4%
17%
2%
1%
1%
Company records, assembled from data from Industrial Market Research Inc. and Information Resources, Inc.
Note: Retail audit estimates may not agree precisely with Brita’s sales records because they are extrapolated from a sampling of retail stores.
17
113
500-024
Exhibit 7
The Brita Products Company
Revenue and Net Income of Brita Systems and Filters, 1998
Total (‘000)
Per Unit
Brita Pitcher Systems
Unit sales
Revenues
Cost of goods sold
Gross margin
Consumer promotion
Feature price reductions
Other trade spending
5,266
$79,800
41,100
38,700
4,000
5,000
7,000
$15.16
7.80
7.36
Brita Filters
Unit sales
Revenues
Cost of goods sold
Gross margin
Consumer promotion
Feature price reductions
Other trade spending
27,413
$112,400
56,200
56,200
1,000
1,000
1,000
$4.10
2.05
2.05
Combined Brita Systems and Filters
Combined revenues
Combined gross margins
Advertising
Combined consumer and trade promotions
Net income before G & A
$192,000
94,900
30,000
19,000
45,900
Source: From company records, modified to preserve confidentiality of margin information but without altering the relative magnitudes of margins.
18
114
UVA-M-0451
Rev. Apr. 12, 2013
THE INDEPENDENT ADVISER FOR VANGUARD INVESTORS
Although Dan Silver’s MBA education was backed by considerable experience in editing and publishing, he felt challenged by some of the details of his new job as director of marketing and operations for the Independent Adviser for Vanguard Investors (TIAVI), a financial newsletter. In particular, many of the complexities surrounding the renting of mailing lists and the direct-mail solicitation of additional customers were new to him. Currently there was growing pressure within his company to cut back on prospecting because several recent efforts had lost money. It was clearly his responsibility to respond to the pressure.
Newsletter Background
Dan Wiener, a former U.S. News & World Report associate editor, started the Vanguard
Adviser in 1991. The newsletter’s purpose was to give advice to investors in the Vanguard Group family of mutual funds and to make money doing so. In 1993, the Vanguard Group was the second-largest group of mutual funds, managing over $100 billion of assets in its more than 70 funds. Approximately 2.5 million individual investors had money invested in one or more
Vanguard funds. For many of these investors, the advice of a Wall Street outsider like Wiener was seen as particularly valuable. Not only was Wiener not associated with Wall Street, he had no ties to the funds about which he gave advice. Because he had nothing to sell, Wiener’s independent advice was seen by subscribers as more valid and trustworthy. In addition, mutualfund companies such as the Vanguard Group were prohibited by law from giving advice on their own funds.
In early 1992, Wiener and his company, the Fund Family Shareholder Association, had brought in two partners from the publishing business to manage the day-to-day operations of the newsletter. Wiener, of course, remained the president. Later that summer, the Fund Family
Shareholder Association hired Silver as director of Marketing and Operations. In May 1993, the newsletter made Marketing Services International, Inc., (MSI) the manager of its member and inquiries file.
This case was prepared by Professor Phillip E. Pfeifer. It was written as a basis for class discussion rather than to illustrate effective or ineffective handling of an administrative situation. Some elements in the case are disguised.
Copyright ¤ 1994 by the University of Virginia Darden School Foundation, Charlottesville, VA. All rights reserved.
To order copies, send an e-mail to sales@dardenbusinesspublishing.com. No part of this publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means— electronic, mechanical, photocopying, recording, or otherwise—without the permission of the Darden School
Foundation. ¸
115
-2-
UVA-M-0451
The newsletter received some early publicity about its feud with the Vanguard Group over the newsletter’s original name, the Vanguard Adviser. Because the Vanguard name was the brand name of the Vanguard family of funds, they at first strongly objected to its use by Wiener as the title of his newsletter. According to their lawsuit, because Wiener had no connection with the Vanguard Group or the Vanguard family of funds, he had no right to call his newsletter the
Vanguard Adviser. The Vanguard Group dropped their lawsuit and made several concessions in
June 1993, in return for the newsletter changing its name to the Independent Adviser for
Vanguard Investors. The publicity surrounding the feud was thought to have contributed to the newsletter’s rapid growth (from 10,000 subscribers to 25,000 within a year). A description of the newsletter can be found in the promotional piece included as Exhibit 1.
Prospecting for New Customers
The newsletter relied heavily on direct-mail “prospecting” to attract new subscribers.1
Slightly over $1 million was spent in 1993 to mail promotional material to carefully selected lists of individuals in the hope that they would purchase an introductory subscription to TIAVI. The promotional package was a regular four-by-nine-inch envelope stuffed with information on the newsletter (see Exhibit 2 for the front and back views of two such envelopes). An initial oneyear membership was offered at a price deeply discounted from its usual price of $139. Sixmonth and two-year memberships were also available. The average amount received with an initial membership was $56. With a one-year membership, subscribers received 12 issues a year of TIAVI, five special reports, and one year’s access to the Fund Family Shareholder Association hotline, a private telephone number offering weekly updates and late-breaking financial news on a 24-hour-a-day basis. As the promotional material stated, “That’s a total value of $307 for only
$99, a savings of $208!”
Silver relied heavily on both MSI and Kim Scott, a free-lance circulation consultant, to select rental lists of individual names and addresses for these prospecting activities. For the most part, the selection of lists was left to the experts who knew that Silver and his partners would carefully monitor and track their results. In general, Silver thought it was a good idea to rent lists through a broker because the seller (renter) of the list paid the broker’s commission. Thus, the broker’s services were “free” to Silver when used to rent lists. In addition, the use of a broker eliminated many of the day-to-day hassles involved in prospecting—something particularly important to a small organization such as Silver’s.
Silver’s decisions, then, were not so much about which lists to use but rather about how aggressive to be. Recently, for the first time, some promotional mailings had failed to breakeven. One line of thinking within the organization was that promotional mailings should “pay for themselves.” In other words, the total cost of the mailing (printing, postage, list rental, etc.) should be recovered from the initial subscription revenues achieved from the mailing. Given that response rates almost always declined when a list was used again and again, it was argued that
1
For the purposes of this case, we will use the terms “subscriber” and “member” interchangeably. Technically, the Fund Family Shareholder Association sold memberships that included a subscription to the newsletter.
116
-3-
UVA-M-0451
once a mailing failed to break even, it made no sense to return to that list. Taken to its logical conclusion, this suggested that a promotional mailing that failed to break-even was also a mistake—the company would have been better off without it.
Exhibit 3 shows the results of the most recent promotional mailing to 10 rental lists.
Included in the last column is total net cash—calculated as subscription revenues minus prospecting costs. Several of the lists showed a loss, and the percentage of lists showing losses had been increasing over time.
Nevertheless, Silver knew he was offering a good product. About 50% of those who tried an initial subscription at the promotional rate converted it to a regular subscription at the end of the year. And the majority (75%) of subscribers renewed in each subsequent year. It cost about
$5 per subscriber to produce and send the promotional material designed to convince them to resubscribe. Because all subscription prices after the initial trial were higher, revenues were higher—about $99 on average. Fulfillment costs (costs to print, insert, and mail the newsletter) were about $12 per year.
At a recent Newsletter Publishers Association conference, Silver had obtained a copy of a formula that could be used to calculate the total number of subscriptions (circulation) to expect from a promotional mailing (see Exhibit 4). At the time, it reminded him of the dividend discount growth formula he had learned in his MBA finance class.
Silver knew he needed to respond to the growing pressure from within the company to do something about the recent prospecting losses. The $64,000 total gain from the recent prospecting effort was not nearly enough to cover his share of company fixed costs. He wondered if there was anything else he could do in addition to developing new and better promotional materials, which was something he worked on all the time. Were the losses something to worry about? Exhibit 5 presents a new set of mailing lists his brokers had recently recommended he try on his next prospecting effort. In the past, he would have mailed promotional material to all of them. But now he was not so sure.
117
-4-
UVA-M--0451
Exhibit
E
1
THE IN
NDEPENDE
ENT ADVIS
SER FOR V
VANGUARD
D INVESTO
ORS
Promotional Materrial
So ource: The Ind dependent Adviiser for Vanguard Investors. U
Used with perm mission. 118
-5-
UVA-M-0451
Exhibit 2
THE INDEPENDENT ADVISER FOR VANGUARD INVESTORS
Example Prospecting Envelopes
119
-6Exhibit 2 (continued)
Example Prospecting Envelopes
Source: The Independent Adviser for Vanguard Investors. Used with permission.
120
UVA-M-0451
-7-
UVA-M-0451
Exhibit 3
THE INDEPENDENT ADVISER FOR VANGUARD INVESTORS
Results of Promotional Mailings
List
List
CPM
$100
$145
$125
$200
$150
$100
$190
$140
$145
$160
Quantity
Mailed
49,525
47,820
39,683
45,920
40,100
41,250
35,644
48,303
38,906
37,783
In-Mail
Cost
$21,8782
$23,598
$20,072
$25,758
$21,451
$19,308
$22,067
$23,498
$20,830
$21,232
Cost per M1
$442
$493
$506
$561
$535
$468
$619
$486
$535
$562
Total
Resp.
501
321
723
340
662
578
902
128
393
526
% Resp.
AAA
BBB
CCC
DDD
FFF
GGG
HHH
JJJ
KKK
LLL
Names
Rented
50,000
50,000
50,000
50,000
50,000
50,000
50,000
50,000
50,000
50,000
1.01%
0.67%
1.82%
0.74%
1.65%
1.40%
2.53%
0.26%
1.01%
1.39%
Net
Cash
$6,1783
($5,622)
$20,416
($6,718)
$15,621
$13,060
$28,445
($16,330)
$1,178
$8,224
Total/Avg
500,000
$145.5
424,934
$219,692
$517
5,074
1.19%
$64,452
ASSUMPTIONS
Activity
Merge/Purge
Printing
Lettershop
Postage
CPM
$30
$242.50
$45
$23
Average Cash Per Order = $56
Source: Created by case writer.
1
Cost per M and CPM refer to cost per thousand.
$21,878 = 50 × ($100 + $30) + 49.525 × ($242.50 + $45 + $23).
3
$6,178 = 501 × ($56) í $21,878.
2
121
-8-
UVA-M-0451
Exhibit 4
THE INDEPENDENT ADVISER FOR VANGUARD INVESTORS
Formula for Total Circulation
Variable
Definition
Circ
Circulation: the total number of subscriptions (including conversions and renewals) resulting from a promotional mailing.
Pcs
Number of promotional pieces mailed.
resp
Response rate to the mailing (a fraction between 0 and 1).
C
Conversion rate: the fraction of responders who are later converted to a regular subscription (a fraction between 0 and 1).
R
Renewal rate: the fraction of regular subscribers who renew (a fraction between 0 and 1).
Circ
§1 C R ·
( Pcs u resp )¨
¸
© 1 R ¹
Source: Created by case writer.
122
-9-
UVA-M-0451
Exhibit 5
THE INDEPENDENT ADVISER FOR VANGUARD INVESTORS
Recommended Mailing Lists for Future Prospecting
List
QQQ
RRR
SSS
TTT
UUU
VVV
WWW
XXX
YYY
ZZZ
Type of List1
Financial
Newsletter
Health
Newsletter
Personal
Finance
Magazine
Business
Magazine
Financial
Newsletter
Political
Newsletter
Books
Tape Series
Financial
Newsletter
Financial
Newsletter
No. of Names
(thousands)
CPM
Average
Purchase2
15
$200
$70
200
$170
$35
500
$150
$30
1,000
$150
$25
75
$210
$129
250
$160
$40
200
50
$90
$90
$20
$99
30
$180
$65
40
$190
$81
Source: Created by case writer.
1
Each of these lists consisted of names of purchasers of a subscription similar to TIAVI. At the request of the company, the exact names of these subscriptions were kept confidential.
2
The average amount the individuals on the list paid to purchase the subscription.
123
124
Bibliography
"Setting Value, Not Price." Leszinski, Ralf; Marn, Michael V. In McKinsey Quarterly, Iss. 1. pp. 98−115. Published February, 1997. McKinsey & Co. Inc. (18 pages).
"Apollo Hospitals: Differentiation through Hospitality." Kulkarni, Suhruta; Makhija, Kripa;
Kumar, Unnikrishnan Dinesh. Case No. IMB425. Published 06/01/2013, Indian Institute of
Management Bangalore, (20 pages).
"Soren Chemical: Why is the New Swimming Pool Product Sinking? (Brief Case)." Rangan,
V. Kasturi; Yong, Sunru. Case No. 4188. Published 04/09/2010, Revised 11/30/2011.
Harvard Business School, (8 pages).
"Note on Behavioral Pricing." Gourville, John. Case No. 9−599−114. Published 05/25/1999,
Harvard Business School Publishing, (12 pages).
"A Practical Guide to Conjoint Analysis." Wilcox, Ronald T. Case No. UVA−M−0675.
Published 08/26/2003, Revised 11/2003. Darden Business Publishing, (9 pages).
"Fidelity Incorporated: Pricing the Fidelity Blue Chip Growth Fund." Wilcox, Ronald T. Case
No. UVA−M−0674. Published 05/20/2003, 2003. Revised 10/08/2002. Darden Business
Publishing, (5 pages).
"Portland Trail Blazers." Wilcox, Ronald T. Case No. UVA−M−0773. Published 4/3/2009,
Revised 2/22/2010. Darden Business Publishing, 2009. (11 pages).
"Customer Profitability and Lifetime Value." Ofek, Elie. Case No. 9−503−019. Published
08/07/2002, Harvard Business School Publishing, (9 pages).
"The Brita Products Co." Deighton, John. Case No. 9−500−024. Published 08/30/1999,
Revised 01/15/2002. Harvard Business School Publishing, (18 pages).
"The Independent Adviser for Vanguard Investors." Pfeifer, Phillip E. Case No.
UVA−M−0451. Published 7/19/1994, Revised 4/12/2013. Darden Business Publishing,
1994. (9 pages).
125
126