Managing Churn By The Numbers

Abstract

Churn (a.k.a. customer turnover) is a significant problem in the telecommunications industry. Given the high cost of acquiring new customers, mining existing databases may provide information to promote customer retention. In everyday language this article presents a method whereby a manager using Microsoft Office is able identify those customers most likely to churn.

The article steps through an example using real world (slightly disguised) cellular phone data. Without using technical language, the reader learns how to generate individual unit profiles and baseline to determine if and when a particular customer is likely to churn.

The process presented is benchmarked using a blind test data sample. The process scored an impressive 80% accuracy in distinguishing active versus disconnected cases.