Market Basket Analysis: Predicting the Next Product Purchase

Student Results:

http://predictive.analyticsight.com/predicting-next-purchase/

Introduction.

EIG offers a portfolio of over 50 products its customers. Unfortunately, every customer is different and a major challenge is to figure out what products are best suited for each customer at a given moment in time. Given customer demographics, time with the company, and other products currently utilized, we must predict what products are most likely to be purchased next by the customer. With that predictive power, EIG would be able to upsell its current clients in an efficient manner.

Assignment Details.

Using SAS Enterprise Miner’s market basket analysis tools, perform the following:

Group 1

  • Identify several hundred association rules
  • Study a subset of those rules that will likely be most relevant to EIG (e.g. rules with substantial lift)
  • Export the subset of rules into a Google spreadsheet and share with the clas

Group 2

  • Identify several hundred sequence rules
  • Study a subset of those rules that will likely be most relevant to EIG (e.g. rules with substantial lift)
  • Export the subset of rules into a Google spreadsheet and share with the class

The Entire Class

  • Explore the various types of reporting tools in Google fusion tables with our hundreds of association and sequence rules
  • Identify which charts, queries, and summaries are most useful for specific business questions