[Solution] Alan Green, an analyst with Pilgrim Bank, needs to prepare - #80214
PILGRIM BANK (B): CUSTOMER RETENTION
Assignment
Alan Green, an analyst with Pilgrim Bank, needs to prepare a report for his boss, Ravi Raman, on whether customers who use on-line banking are more profitable, and if adoption of on-line banking by customers makes them more profitable. To prepare analyses to support his report, Alan has obtained data for a random sample of 5,000 customers for both 1999 and 2000. Alan has discovered that there was not a statistically significant difference in mean customer profitability for on-line and off-line customers in 1999, but has a feeling there may be more information that could help Raman decide how to manage customer profitability. To help him determine what that information is, Alan has asked for your help in answering these questions.
a) Why is there a need to predict future customer profitability? In other words, if Pilgrim Bank knows how profitable a customer is today, why do they need to worry about how profitable that customer is tomorrow?
b) Is there a significant difference in 2000 profitability for customers for whom age and/or income data is missing and for customers for whom that data is not missing? Should age and/or income be included as independent variable(s) in a multiple regression model to predict customer profitability in 2000, and why or why not? HINT: This question cannot be answered using regression; you need to use a different tool.
c) Does geographic region predict customer profitability in 2000? Should geographic region be included as an independent variable in a multiple regression model to predict customer profitability in 2000, and why or why not?
d) Does tenure with the bank predict customer profitability in 2000? Should tenure be included as an independent variable in a multiple regression model to predict customer profitability in 2000, and why or why not?
e) How well does customer profitability in 1999 predict customer profitability in 2000? Should customer profitability in 1999 be included as an independent variable in a multiple regression model to predict customer profitability in 2000, and why or why not?
f) How well does the use of on-line banking in 1999 predict customer profitability in 2000? Should on-line banking be included as an independent variable in a multiple regression model to predict customer profitability in 2000, and why or why not?
g) Are there any other variables included in the data set that can help predict customer profitability in 2000?
h) What is the best regression model for predicting customer profitability in 2000, and why?
i) How many customers in Alan’s sample left the bank from 1999 to 2000?
j) What implications do all of these analyses have for the strategies Pilgrim Bank might consider using to manage customer profitability? Write a memo to Alan that provides information to help answer these questions, and include relevant analyses as exhibits (at a minimum include an exhibit with the regression model identified as the best). Be sure to specifically answer questions a) and j) in your memo and describe to Alan how you decided what variables to include in the regression model you found to be the best.
Deliverable: Word Document
and pdf
