Thursday, May 12, 2011

Prepaid Customer Churn Prediction

To survive competition, telecommunications service providers must detect the main reasons for both expected churn and the churn that happens after an event has taken place in prepaid category because this information can help them to customize their offers.

It can be a tool to effectively anticipate the demands of their key customers who have the highest churn propensity, fully knowing that retention can have a huge impact on the lifetime value (LTV) of a customer.

A study has recently been undertaken and contributed to Inside Revenue Management by Sanket Jain, an Advisory Consultant at IBM India. The objective of this study is to use analytic tools, in this case IBM’s SPSS, with synthetic data for doing prediction and later to analyze the churning and non-churning customers.

The advent of Mobile Number Portability (MNP) in India and its impact on the rate of churn is also discussed. With all the events like MNP that are going in this dynamic field of pay-as-you-go phones, it is important to prepare a base model using analytic tools like SPSS and then to enlist additional input variables that can increase the model accuracy.

Once a robust model is built, it can be further used to explore directions for customer retention. In the end, a literature review, together with some ideas of that of the author, has been done of the promotions that can be designed by service providers with their prepaid subscribers in mind.


http://www.tmforum.org/browse.aspx?linkID=45304&docID=14871

No comments:

Post a Comment