On Analyzing Churn Prediction in Mobile Games

TL;DR

This paper introduces a highly accurate churn prediction method for mobile games that considers individual user usage periods, achieving 96.6% accuracy and improving existing algorithms.

Contribution

The paper presents a novel churn prediction approach that accounts for individual user usage periods, enhancing accuracy over existing methods.

Findings

Achieved 96.6% churn prediction accuracy on real data.

Improved existing churn prediction algorithms with the new method.

Demonstrated the importance of individual user usage periods in prediction.

Abstract

In subscription-based businesses, the churn rate refers to the percentage of customers who discontinue their subscriptions within a given time period. Particularly, in the mobile games industry, the churn rate is often pronounced due to the high competition and cost in customer acquisition; therefore, the process of minimizing the churn rate is crucial. This needs churn prediction, predicting users who will be churning within a given time period. Accurate churn prediction can enable the businesses to devise and engage strategic remediations to maintain a low churn rate. The paper presents our highly accurate churn prediction method. We designed this method to take into account each individual user's distinct usage period in churn prediction. As presented in the paper, this approach was able to achieve 96.6% churn prediction accuracy on a real game business. In addition, the paper shows that other existing churn prediction algorithms are improved in prediction accuracy when this method is applied.