Predicting lapsing customers with logistic regression approach in retail

Data mining is the process of exploring meaningful information in big and complex data to create a valuable business strategy. One of the major application fields of data mining is, predicting customers having a tendency of disconnecting from the services of the company. Also called as churn analysis, it provides predictive information to the companies to score customers having churn risk and then enables them developing retention strategies such as targeted campaigns. This study is executed by using transaction data of customers enrolled in loyalty programme which belongs to a multinational retail company operating in Turkey. It is aimed to score customers having a churn tendency in next 13 weeks and then helping to develope retention strategies based on these scores. In order to explore churn customer profiles , Factor Analysis and Logistic Regression methods are applied and results of the application is presented.