Module 22: Revitalizing Dell
1
Last lecture
• Home Depot revenue (forecasting)
• Using correlation to choose lag • Using Durbin-Watson statistic to test missing drivers • Out-of-sample model validation
2
Dell’s success strategies
• Direct model (marketing)
– “Cut out the middlemen.” – NC born Harlem drug lord Frank Lucas
• Mass customization (design)
– Modularity – Component commonality – Postponement
• Lean manufacturing (operations)
– Just-in-time inventory management – Virtual integration
3
Dell’s challenges
• Direct channel does not cover all markets; competitors attempt to copy direct model • Commoditization of PCs (except Apple) • De-emphasis of hardware: SaaS (Software as a Service) and cloud computing • Mobile revolution (iPhone, 2007)
4
Revitalizing Dell
• 01/2007: Michael Dell reassumed role of CEO • 05/2007: Dell began selling PCs at Wal-Mart • 07/2007: Dell launched a line of bright-colored laptops • 09/2009: Dell acquired Perot Systems for $3.9B
• What does the future hold for Dell?
5
Revitalizing Dell: Forecast Dell’s 2009 and 2010 revenues • Work through the “Proposed Steps” of Case 9-1 Revitalizing Dell in your textbook
– Make lagged drivers – Use correlation to pick a lagged driver – Build a linear forecast model using regression, perform DW test on residuals – Repeat if residuals do not pass DW test
• Forecast revenues and generate 95% prediction intervals for 2009 and 2010
6
Revitalizing Dell: Bright forecast
7
Revitalizing Dell: Harsh reality
8
Revitalizing Dell: What did our model miss? • HP sales data not “clean” (starts to include Compaq since 2002) • Ratings lack variation (could have used Dell’s rating – HP’s rating) • Up-and-coming competitors (Acer, Lenovo) • Mobile revolution: iPhone (2007), Android phones (2008), iPad (2010) • Financial crisis and recession (2008)
9
Using past to forecast future
“The best prophet of the