CMF: A framework to improve the management of customer churn

Acquiring new customers in any business is much more expensive than trying to keep the existing ones. Many churn management models have been developed over the years, mostly focusing on prediction accuracy and not considering the range of parameters and processes essential to manage the system as a whole. This study presents CMF (churn management framework), a framework which tries to covers most of the requirement of a churn management system. Some of the results indicate that the analysis of churn reasons can have a positive impact on system efficiency. Also the preprocessing requirements of such model differ from other prediction systems thus requiring close attention. In conclusion, the profit gain of a company applying CMF is proved to be much higher than the profit gain from an ordinary churn prediction program.