Identifying the Vips at Etots
2 Pages Posted: 30 May 2017
Abstract
This case contains the instructions for a hands-on data-mining exercise. The purpose of the exercise is to identify high value customers of an online retailer in order to give them VIP treatment. The exercise uses purchase data for 60,000 customers. The background is described in "Customer Care at eTots.com" (UVA-M-0649).
Excerpt
UVA-M-0650
Rev. Mar. 25, 2014
IDENTIFYING THE VIPS AT ETOTS
In this exercise, you will be asked to “identify” the VIP customers in the test set of 30,000 customers. Data for the test set of customers are found in the “VIP Exercise Decision Set” file (UVA-M-0649X2). Remember that the customers in the test set are identical to the customers in the training set (UVA-M-0649X1) in almost every way. The test group contains 10,000 customers who purchased during Holiday 2000 and 20,000 who did not. The test set and training set are split samples of an original set of 60,000 customers.
We know the Holiday 2000 purchase results for the new group of customers but have not included them in the file “VIP Exercise Decision Set” (UVA-M-0649X2). The two held-out variables (Holiday 2000 repeat and Holiday 2000 MVA) are found in the faculty spreadsheet file “VIP Exercise Secret Answers” (UVA-M-0649TNX), which your instructors will use to evaluate and score your team's performance on this exercise.
You must decide which of the 30,000 customers in the test group to designate for VIP treatment. You will submit your decisions in an Excel file containing two columns. The first column should contain the 30,000 customer IDs. The second column should contain a 0/1 variable, where 1 represents VIP treatment and 0 represents regular treatment.
. . .
Keywords: data analysis, marketing management
Suggested Citation: Suggested Citation