Predicting customer behavior in telecommunications

The wireless service subscriber calls a customer service representative to complain about dropped calls. During the conversation with the customer, the CSR views a display that shows this customer's probability of churn-switching from this service provider to another-as well as the most probable reasons to churn and the best strategy to retain this customer. The CSR then quickly responds to the subscriber according to the system's recommendation. This is an intelligent customer-care system designed to predict customer behavior. Predicting customer churn is a component in the decision framework for retaining customers and maximizing profitability. Companies can use these probability and revenue estimates in a decision-theoretic framework to determine a churn intervention strategy and a profitability optimization strategy. Predicting customer behavior helps service providers build customer loyalty and maximize profitability. For the success of a project, data preparation is often a critical part of the predictive algorithm.