Customer Churn Prediction in the Online Gambling Industry: The Beneficial Effect of Ensemble Learning

K. Coussement and Koen De Bock ()
Additional contact information
K. Coussement: LEM - Lille - Economie et Management - Université de Lille, Sciences et Technologies - CNRS - Centre National de la Recherche Scientifique

Post-Print from HAL

Abstract: The online gambling industry is one of the most revenue generating branches of the entertainment business, resulting in fierce competition and saturated markets. Therefore it is essential to efficiently retain gamblers. Churn prediction is a promising new alternative in customer relationship management (CRM) to analyze customer retention. It is the process of identifying gamblers with a high probability to leave the company based on their past behavior. This study investigates whether churn prediction is a valuable option in the CRM palette of the online gambling companies. Using real-life data of poker players at bwin, single algorithms, CART decision trees and generalized additive models are benchmarked to their ensemble counterparts, random forests and GAMens. The results show that churn prediction is a valuable strategy to identify and profile those customers at risk. Furthermore, the performance of the ensembles is more robust and better than the single models. (

Date: 2013
References: Add references at CitEc
Citations: View citations in EconPapers (27)

Published in Journal of Business Research, 2013, 66 (9), pp.1629-1636. ⟨10.1016/j.jbusres.2012.12.008⟩

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
Journal Article: Customer churn prediction in the online gambling industry: The beneficial effect of ensemble learning (2013)
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:hal:journl:hal-00788063

DOI: 10.1016/j.jbusres.2012.12.008

Access Statistics for this paper

More papers in Post-Print from HAL
Bibliographic data for series maintained by CCSD ().