Customer Churn Prediction in Chinese Traditional Broadcasting Industry: A Positive Analysis

In order to better respond to market competition, national broadcast service providers such as cable network enterprises have to recognize customer preference and forecast customer churn intention before their rivals do so. In this paper, data recorded by the digital set-top box and consuming records in cable network's finance system is used for analyzing customer churn. First, the results of this research indicate that customer watching intensity, customer consumption amount and customer paying habits have significant influence on customer churn. And the effect of customer watching intensity on customer churn is moderated by customer watching preference. Second, this research uses logistic regression to build the prediction model. Finally, we propose different user retention strategies targeting the churn customers. This study demonstrates that our proposed user churn factors are efficient and capable of providing effective marketing strategies for cable network enterprises.