What advantages are there in using this approach to calculate customer lifetime value and what reservations would you have with using it exclusively?
To estimate customer lifetime value, CBR Bank uses a probability transition matrix for two bank products: car loans (CL) and credit cards (CC). The number in each cell is the probability that a customer who had the product in that row in one year will have product indicated in the column in the next year.
CL only CC only Both Neither
CL only 0.2 0.1 0.5 0.2
CC only 0.1 0.5 0.2 0.2
Both 0.1 0.1 0.7 0.1
Neither ~0.0 ~0.0 ~0.0 ~1.0
The expected profit for the bank from car loans is ($100) per year, and for credit cards it is $1,000 per year.
What does the probably matrix tell us about CBR’s expectations since, after all, they developed this matrix for a specific purpose? For a customer who has only the car loan in the current year (year 1), what is the expected profit generated for the bank by this customer next year (year 2)? What would the expected profits be in year 3? What is the present value of the expected profits during each of the three years, and in total, using an 8% discount rate (contrary to what we did in class, discount the first year)? What advantages are there in using this approach to calculate customer lifetime value and what reservations would you have with using it exclusively?