Manzana Insurance

1. How is the Fruitvale branch doing? On viewing the financial performance of Manzana Ltd., we see that the branch profitability is decreasing. There is a huge backlog of policies, which is resulting in delays for current orders too. The number of new policies and endorsements appeared to be stagnating, whereas the rest of the industry reported moderate growth rates. According to the figures of the present quarter, Fruitvale had shown a decline in insuring new policies whereas Golden Gate had issued 100 more policies than them in the same quarter. Moreover, the turnaround time, the time from policy request to policy issue, for Fruitvale was 5 days, which was, much more than the turnaround time of Golden Gate (2 days). The branch was also losing on renewals, which amounted to nearly 33%, again more than double of Golden Gate. The renewal losses represent a significant loss of business and an overall reduction in the number of policies in force. While new policies have not increased substantially, renewal policies are being lost, thus, effective growth isn't happening. 2. What are the causes of these problems? One of the causes is the wrong method used in calculating the turnaround time. Their calculations show a much higher time than is actually the case, hence, agents don't want to work for them and readily switch to the competition. Another cause is the employees' lack of willingness to work more at RERUNs and RAINs. When the conversion rate of RAP to RUN is only 15%, it seems that a lot of time is being wasted. The revenue figures for RAPs hasn't been given explicitly, but still, a proper cost-benefit analysis needs to be done to establish priorities. The employees aren't working enough on renewals, hence, they are losing their existing customers. This can be partly traced to the practice of working on RUNs and RANs first. Another part of the problem is that in their quest for up-to-date information, the relevant data are released just one day before the date of renewal. When the information is released, it is not acted quickly upon as the employees focus more on RUNs and RAPs. Elementary marketing theory states that it costs at least 5 times to attract a new customer than to please an existing customer. Hence, their practice isn't cost-effective, which is being reflected in the falling premiums. Another problem is that agents' contracts aren't being renewed at the same pace as that of the competition. Hence, these agents were recommending other insurance providers. We can look at the utilization percentage of the groups of employees. For example, let's consider distribution. Weighted average time = 41 min. Arrival rate = 39 policies/day (given) Total time available = 7.5 * 60 * 4 = 1800 min/ day Capacity = 1800 / 41 = 43.9 policies. Hence, utilization = 39 / 43.9 = 89% Similarly, if we calculate the % utilization for all groups, we get the following:

Distribution 89%
Underwriters 82%
Rating 76%
Policy Writing 64%

Hence, we see that the rating and policy writing departments have a lot of idle time.
This is a symptom, the cause lies in incorrect calculation of turnaround time and doing the work in series rather than in parallel.

3. Can you identify the problems in the way Manzana is calculating turnaround time in exhibit 3? In calculating the turnaround time, Manzana takes into account the 95th percentile of the completion time. This is a very conservative estimate, the calculation should be done using the mean. This results in a difference by a factor of 2-3, which has a significant effect on the turnaround time. Another issue is that they assume a series operation while calculating the turnaround time. Ideally, work should be done simultaneously by the different groups of people. When these 2 factors are taken into consideration, the actual turnaround time will be reduced substantially. Operating Steps RUNs RAPs RAINs RERUNs Throughput days
Distribution
Pending 1 3 1 11
Av. per clerk 0.25 0.75 0.25 2.75
Mean time 68.5 50 43.5 28
Total time 17.125 37.5 10.875 77 0.3167 Underwriting
From above 1 3 1 11
Pending 1.82 0.092 3.11 23.65
Total 2.82 3.092 4.11 34.65
Av. Per team 0.94 1.03 1.37 11.55
Mean time 43.6 38 22.6 18.7
Total time 41 39.14 30.96 216 0.727
Time elapsed 58.125 76.64 41.835 293
Rating
From above 2.82 3.092 4.11 34.65
Pending 0.23 0.815 0.36 3.86
Total 3.05 3.9 4.47 38.51
Av. Per clerk 0.38 0.4875 0.56 4.81
Mean time 75.5 64.7 65.5 75.5
Total time 28.69 31.54 36.68 363.155 1.02
Time elapsed 86.815 108.18 78.515 656.155
Policy writing
From above 3.05 3.9 4.47 38.51
Pending 0 NA 0 0
Total 3.05 4.47 38.51
Av. per person 0.61 0.894 7.7
Mean time 71 54 50.1
Total time 43.31 48.276 385.77 1.06
Total TAT 3.12

Explanation of calculation of pending work:
A sample calculation of the pending work for underwriters in RUNs is shown. The same method has been followed for calculating the pending work at all stages.
Mean time for a policy = 43.6 min.
Total time elapsed = 17.125 min.
In this time, work done by 3 groups = 3 * (17.125/43.6) = 1.18 policies.
Pending work = 3 – 1.18 = 1.82 policies.

Hence, while calculating TAT every Monday morning, initially, the system won't be in steady state. The above method should be used for calculating TAT.
As work progresses, the system achieves a steady state. At steady state, the policy writers won't be waiting for the raters to finish their work. Hence, turnaround time at steady state can be calculated by using Little's law.
WIP = 82, throughput rate = 40 arrivals/day, hence,
TAT ~ 2 days.

Hence, we see that the turnaround time can be reduced substantially, which is a significant reduction. The problem with the initial method of calculation was that it allowed for a lot of idle time to employees in the subsequent stages. Hence, the employees in the rating and policy writing stages were under-utilized. With the above method of working simultaneously, the pending queue becomes shorter, resulting in quicker turnaround times. This will definitely contribute to increasing the business of Manzana Insurance Ltd.

4. What recommendations would you make to Fruitvale?
• One recommendation is to faithfully follow the policy of first-in-first-out. As we have seen that the current policy is making losses, it would serve better to retain old customers.
• The above method should be used for calculating turnaround time, so that policies can be issued quickly and backlog cleared. It will give a competitive edge to Manzana.
• RERUNs should be given more importance than RAPs as the conversion ratio of RAPs to RUNs is only 15%. The extra revenues that RAPs may generate doesn't seem to compensate for the fall in premiums due to loss of old customers.
• The information for RERUNs should be released earlier so that renewal losses can be minimized. An alternative is to form a dedicated team for serving only RERUNs so that renewal losses are reduced.
• Contracts with agents should be renewed periodically so that they don't switch to the competition.
• From the data given in Exhibit 4, the following table can be constructed for the coefficient of variation:

Operating Steps RUNs RAPs RAINs RERUNs
Distribution 44.8% 49.8% 21.1% 22.1%
Underwriting 73.4% 64.5% 51.7% 105.9%
Rating 27.1% 21% 24.3% 12.8%
Policy Writing 14.5% 15.9% 19% From the above table, we see that there are 6 processes where the coefficient of variation is above 40%. This means that a lot of variability is present in the system. Hence, Lq increases. What can be done in this case is that requests can be classified on the basis of most likely time taken. They can be handled in separate queues so as to reduce the service time. This will result in shorter queues.

• The 4 teams are unevenly utilized as has been shown before. Hence, the utilization across teams can be balanced by shifting an employee from the rating or writing department to the distribution or underwriting department.