Practical Predictive Analytics

Clustering example using k-means

This example will illustrate a simple RFM clustering example. RFM analysis is simple database marketing that can be used to segment out your best (and worst) customers. It is a good example to use for demonstrating cluster analysis since there are only three variables, and can be segmented, theoretically, into a much larger number of marketing segments.

The data comes from the CDNOW database, which contains the entire purchase history up to the end of June 1998 of the cohort of 23,570 individuals who made their first-ever purchase at CDNOW in the first quarter of 1997 (Fader, 2001).

Here are three variables that we will be using:

• Date of purchase: This is the date that the item was purchased and will be used ...