ABSTRACT
The telecommunications industry with an approximate annual churn rate of 30% can nowadays be considered as one of the top sectors on the list of those suffering from customer churn. Although different studies have focused on developing a predictive model for customer churn under contractual settings, the mobile telecommunications industry, performing in a non-contractual setting in which customer churn is not easy to define and trace, has always been neglected in such investigations. In this study, we have developed a dual-step computer-assisted model in which a clustering model and a classification model are employed for defining and predicting customer churn. Results indicate the promising performance of the proposed models in identifying future churners.
