Forecasting customer lifetime value: a statistical approach

Erniel B. Barrios, Joseph Ryan G. Lansangan

Research output: Contribution to journalArticleResearch

Abstract

We propose a method of forecasting customer lifetime value using the customer usage database, sampling strategy, segmentation, modeling, and validation techniques in data mining. The highly heterogeneous customer database being mined will allow the inclusion of uncertainty components in the estimation of customer lifetime value. A hazard function model based on a truncated lifetime data of customers can provide adequate information to compute the value of the incentive to be offered and the length of the lock up period for the customer.
Original languageEnglish
Pages (from-to)23-34
Number of pages12
JournalPhilippine Management Review
Volume19
Publication statusPublished - 2012
Externally publishedYes

Keywords

  • data mining
  • customer lifetime value
  • customer relationship management
  • truncated data
  • hazard function model

Cite this