Churn prediction in telecommunication industry in Kenya using decision tree – Case study orange Kenya

Abstract

Customer churn in the telecommunication industry is still a big problem because emerging new technologies, lower costs, among other factors. Right now Kenya is developing at a very fast rate and the market is getting new players leaving the only way to gain customers is by winning them over from the competitors. The retention of customers is becoming a huge challenge and the cost of acquiring new customers is more expensive, therefore by collecting information already available to the telecom industry can go a long way to helping them. The best way for the telecommunication industries to address this is to develop precise and reliable predictive models so as to identify potential churners by understanding their behavior and trends beforehand so as to introduce them to the programs suited to their needs in a bid to retain them. Orange Kenya was used as a case study to develop a predictive a model in an aim to reduce the customer churn rate. The objective of the study was to find the extent of the customer churn in Orange Kenya and make it obligatory for the telecommunication industry to do a churn prediction analysis. Also use the available resources to design a model for customer churn prediction for the pre-paid users. Questionnaires use was to find out customer’s perspective. The project provides a framework for churn prediction model and implemented using data mining. The final results show the prediction rate for the model used.