Customer Lifetime Value Prediction Model

Abstract

In the services sector and especially in telecommunications sector gauging the future value of new customer acquisitions and effectiveness of marketing campaigns has always been a challenge. Mobile operators invest significant percentage of their revenues in advertising and customer acquisition; therefore it is pertinent that the return on this investment can be measured objectively. Due to focus on short term results it is observed that subscriber’s loyalty towards network has decreased, increasing churn and in the long run hurting profitability. A model which can predict the value of each subscriber therefore, would be immensely beneficial in gauging the quality of new subscribers and help in valuation of revenue generation potential of the existing subscriber base. Keeping this in view, telecom variables e.g. Calls, SMS, Internet usage, Value Added Services usage and their respective parameters e.g. call duration, count, frequency, session time, price plan and bundle type were selected for this study, Customer life time value (CLV) was calculated based on two months data by using SAS (Statistical Analysis System) and Teradata. Customers were divided into different segments based on their CLV. Customers were subsequently targeted by customized promotional campaigns, and post campaign, their CLV were calculated. Hence, the improvement and effectiveness of tailored campaigns was gauged. Companies will also be able to reduce churn by directing their efforts to those subscribers who can be persuaded to stay with network and rectifying their exact needs which would have caused them to detract from the network in the future.