Murdoch University Research Repository

Welcome to the Murdoch University Research Repository

The Murdoch University Research Repository is an open access digital collection of research
created by Murdoch University staff, researchers and postgraduate students.

Learn more

Identifying customers likely to churn

Zorn, S., Jarvis, W. and Bellman, S. (2008) Identifying customers likely to churn. In: ANZMAC 2008: Marketing: Shifting the Focus from Mainstream to Offbeat, 1 - 3 December 2008, Olympic Park, Sydney

PDF - Published Version
Download (135kB)


As acquiring new customers is costly, it seems logical to keep and satisfy long-time customers rather than to acquire new customers. To reduce churn rates, firms should manage customers proactively to avoid losing churned customers. The study investigated how an Australian DVD rental firm can use customer data to derive indicators of satisfaction, attitude, and commitment to improve the prediction of customer churn in comparison to models calibrated on purchasing behaviour alone. The most significant predictor of churn in these data was a measure of uncertainty and commitment: the number of times a customer changed their subscription plan.

Item Type: Conference Paper
Murdoch Affiliation(s): Interactive Television Research Institute
Conference Website:
Item Control Page Item Control Page


Downloads per month over past year