Churn; Decision Tree; Data Mining; iGaming; Online Gaming; Customer relationship management; Marketing


Gaming and Casino Operations Management

Document Type

Original Research Article


With the potential expansion of legalized online gaming in the United States as well as in the global market, customer retention is critical to the continued growth and success of an online casino. While customer churn prediction can be an essential part of customer retention efforts, it has received very little attention in the gaming literature. Using historical online gaming data, this study examines whether player churn (attrition) can be predicted through an application of a decision tree data mining algorithm called Exhaustive CHAID (E-CHAID). The results of this empirical study suggest that the predictive model based on the E-CHAID method can be a valuable tool for identifying potential churners and understanding their churn behavior. Additionally, this study shows how the classification rules and propensity scores extracted from a decision tree churn model can be used to identify players at risk of churn. The patron play and visitation parameters that are closely associated with churn are also discussed. This study contributes to the gaming literature by focusing on online players’ churn prediction through a data-driven approach. Finally, it discusses proactive approaches for churn prevention.