Predicting subscriber dissatisfaction and improving retention in the wireless telecommunications industry

Competition in the wireless telecommunications industry is fierce. To maint ain profitability, wireless carriers must control churn, which is the loss of subscribers who switch from one carrier to another, We explore technique s from statistical machine learning to predict churn and, based on these pr edictions, to determine what incentives should be offered to subscribers to improve retention and maximize profitability to the carrier. The technique s include legit regression, decision trees, neural networks, and boosting. Our experiments are based on a database of nearly 47 000 U.S. domestic subs cribers and includes information about their usage, billing, credit, applic ation, and complaint history. Our experiments show that under a wide variet y of assumptions concerning the cost of intervention and the retention rate resulting from intervention, using predictive techniques to identify poten tial churners and offering incentives can yield significant savings to a ca rrier. We also show the importance of a data representation crafted by doma in experts. Finally, we report on a real-world test of the techniques that validate our simulation experiments.