using neural networks for customer churn prediction ...

Proceedings del X° Convegno Annuale della Società Italiana Marketing: “Smart Life. Dall’Innovazione Tecnologica al Mercato”, 3-4 Ottobre 2013, Università degli Studi di Milano-Bicocca, Milano, pp. 1-13, ISBN: 978-88-907662-1-3

USING NEURAL NETWORKS FOR CUSTOMER CHURN PREDICTION MODELING: PRELIMINARY FINDINGS FROM THE ITALIAN ELECTRICITY INDUSTRY* ABSTRACT Having an accurate customer churn prediction model become a very important issue for each firm to indivduate customers who are truly at risk to churn and retain loyal customers. Our study applies neural networks (NNs) to a sample of 1.463.199 indirect customers of an Italian electricity firm. Results show that (1) when compared with a traditional method, NNs can make a better contribution to churn prediction, and 2) when NNs are separately calculated according to age, prediction accuracy is markedly higher. Keywords: Churn Management, Prediction Model, Neural Network, Decision Tree

1) INTRODUCTION The electricity industry is converging toward a competitive framework where a market environment is replacing the traditional monopolistic scenery (Catãlao et al., 2007). In the regulated framework the electricity industry’s attention mainly focused on load forecasting, having little need for tools hedging against churn risk given the deterministic nature of electricity prices. In nowadays, the deregulation of the electricy market has increased the competition among companies and turned electricity into a traded commodity to be sold and bought at market prices. Consequently, customer churn has received ever more attentions in this industry. It amounts to approximately 17% in Italy, 12% in UK and 5% in Germany (Busacca, 2004). Marketing literature has stated that it is more costly to engage a new customer than to retain an existing loyal customer because: (1) attracting new clients costs more than retaining existing customers (Athanassopoulos, 2000; Bhattacharya, 1998; Colgate et al., 1996, 2000; Verbeke et al., 2011, 2012); (2) long-term customers generate higher profits, tend to be less sensitive to competitive marketing activities, become less costly to serve, and may provide new referrals through positive word-ofmouth, while dissatisfied customers might spread negative word-of mouth (Ganesh et al., 2000; Mizerski, 1982; Paulin et al., 1998; Stum and Thiry, 1991; Verbeke et al., 2011; 2012; Zeithaml et al., 1996); (3) losing customers leads to opportunity costs because of reduced sales (Rust and Zahorik, 1993). A small improvement in customer

* Niccolò Gordini, Assistant Professor of Management, University of Milano-Bicocca, [email protected] Valerio Veglio, Fellow Researcher, University of Milano-Bicocca, [email protected]

1


retention can therefore lead to a significant increase in profit (Van den Poel and Lariviere, 2004), whilst churn may have damaging consequences for the companies. Customer churn is a term used to indicate the customer movement from one company to another. A way to manage customer chum is to predict which customers are most likely to chum and then target incentives to those customers to induce them to stay and to limit their losses (Bolton et al., 2000; Ganesh et al., 2000; Lemmens and Croux, 2006; Shaffer and Zhang, 2002). This approach enables the firm to focus its efforts on customers who are truly at risk to chum, and it potentially saves money that would be wasted in providing incentives to customers who do not need them (Neslin et al., 2006). As churn management is an important activity for companies to retain loyal customers, the ability to correctly predict customer churn is necessary. In this paper, we test the forecasting capability of neural networks (NNs) in developing a customer churn prediction model for a sample of 1.463.199 customers of an Italian electricity firm using 14 variables. We, then, compare this capability to that of decision tree (DT) and we investigate how the structure of the sample in terms of custormers’age influences the accuracy of the model. The remainder of the paper is structered as follows. The next section provides a brief review of literature based on customer churn prediction modeling. In sections 3 and 4 the hypotheses and methodology of our study are briefly described, whereas in the fifth section the sample and the variables are described. Following, the section discusses the empirical results of our study. The last section concludes the paper with some managerial implications, limitations, and future research avenues.

2


REFERENCES Athanassopoulos A.D. (2000). Customer Satisfaction Cues to Support Market Segmentation and Explain Switching Behavior. Journal of Business Research, 47 (3), pp. 191-207. Au W., Chan K. and Yao X. (2003). A Novel Evolutionary Data Mining Algorithm with Applications to Churn Prediction. IEEE Transactions on Evolutionary Computation, 7(6), pp. 532–545. Baesens B., Viaene S., Van den Poel D., Venthienen J. and Dedene G. (2002). Bayesian Neural Network Learning for Repeat Purchase Modeling in Direct Marketing. European Journal of Operational Research, 138(1), pp. 191-211. Bhattacharya C.B. (1998). When Customers Are Members: Customer Retention in Paid Membership Contexts, Journal of the Academy of Marketing Science, 26(1), pp. 31-44. Bolton R.N., Kannan P.K. and Bramlett M.D. (2000). Implications of Loyalty Program Membership and Service Experiences for Customer Retention and Value, Journal of the Academy of Marketing Science, 28(1), pp. 95-108. Buckinx W. and Van den Poel D. (2005). Customer Base Analysis: Partial Defection of Behaviourally Loyal Clients in a Non-contractual FMCG Retail Settingg. European Journal of Operational Research, 164(1), pp. 252–68. Burez J. and Van den Poel D. (2007). CRM at a Pay-TV Company: Using Analytical Models to Reduce Customer Attrition by Targeted Marketing for Subscription Services. Expert System with Applications, 32, pp. 277-288. Burez J. and Van den Poel D. (2009). Handling Class Imbalance in Customer Churn Prediction. Expert Systems with Applications, 36(3), pp. 4626–4636. Busacca B. (2004). Energia: La Liberalizzazione Avanza e anche i Piccoli Clienti Cominciano a Scegliere, www.busacca.it. Catãlao J.P.S., Mariano S.J.P.S., Mendes, V.M.F. and Ferreira, L.A.F.M. (2007). Short-Term Electricity Prices Forecasting in a Competitive Market: A Neural Network Approach. Electric Power Systems Research, 77, pp. 1297–1304. Ciampi F. and Gordini N. (2013). Small Enterprise Default Prediction Modeling through Artificial Neural Networks: An Empirical Analysis of Italian Small Enterprises. Journal of Small Business Management, Vol.51(1), pp. 23-45. Colgate M., Stewart K. and Kinsella R. (1996). Customer defection: A study of the Student Market in Ireland. International Journal of Bank Marketing. 14(3), pp. 23– 29. Colgate M.R. and Danaher P.J. (2000). Implementing a Customer Relationship Strategy: The Asymmetric Impact of Poor Versus Excellent Execution. Journal of the Academy of Marketing Science, 28(3), pp. 375-87. Coussement K. and Van den Poel D. (2008). Churn Prediction in Subscription

3


Services: An Application of Support Vector Machines while Comparing Two Parameter-Selection Techniques. Expert Systems with Applications, 34(1), pp. 313– 27. Datta P., Masand B., Mani D. and Li B. (2000). Automated Cellular Modeling and Prediction on a Large Scale. Artificial Intelligence Review, 14, pp. 485–502. Demyanyk Y. and Hasan I. (2010). Financial Crises and Bank Failure: A Review of Prediction Methods, Omega, 38, pp. 315–324. Eiben A., Koudijs A. and Slisser F. (1998). Genetic Modeling of Customer Retention. Lecture Notes in Computer Science, 1391, pp. 178–186. Ganesh J., Arnold M.J. and Reynolds K.E. (2000). Understanding the Customer Base of Service Providers: An Examination of the Differences Between Switchers and Stayers. Journal of Marketing, 64 (July), pp. 65-87. Giudici P. (2003). Applied Data Mining. Chichester, UK: Wiley. Giudici P. (2010). Data Mining. Milano, Italy: McGraw-Hill. Giudici P. and Figini S. (2009). Applied Data Mining for Business and Industry. New York, USA: John Wiley & Sons, Inc. Gordini N. and Veglio V. (2013). Customer Relationship Management and Data Mining: A Classification Decision Tree to Predict Customer Purchasing Behavior in Global Market, in Vasant, P. (Ed.), Handbook of Research on Novel Soft Computing Intelligent Algorithms: Theory and Practical Applications, Hershey, PA, USA: IGIGlobal, pp.1-40. Ghosh A., Neslin S.A. and Shoemaker R.W. (1984). A Comparison of Market Share Models and Estimation Procedures. Journal of Marketing Research, 21 (May), pp. 202-210 Guido G., Prete M.I., Miraglia S. and De Marec I. (2011). Targeting Direct Marketing Campaigns by Neural Networks. Journal of Marketing Management, 27(9–10), pp. 992–1006. Ha K., Cho S. and MacLachlan D. (2005). Response Models Based on Bagging Neural Networks. Journal of Interactive Marketing, 19(1), pp. 17–30 Haughton D. and Oulabi S. (1993). Direct Marketing Modeling with CART and CHAID. Journal of Direct Marketing, 11(4), pp. 42–52. Hornik K. (1991). Approximation Capabilities of Multilayer Feed Forward Networks, Neural Networks, 4(2), pp. 251–257. Hung S., Yen D. and Wang H. (2006). Applying Data Mining to Telecom Churn Management. Expert Systems with Applications, 31, pp. 515–524. Hwang H., Jung T. and Suh E. (2004). An LTV Model and Customer Segmentation Based on Customer Value: A Case Study on the Wireless Telecommunication Industry. Expert Systems with Applications, 26(2), pp. 181–8.

4


Kass G.V. (1980). An Exploratory Technique for Investigating Large Quantities of Categorical Data. Applied Statistics, 29(2), pp.119-127. Kumar A., Rao V.R. and Soni H. (1995). An Empirical Comparison of Neural Network and Logistic Regression Models. Marketing Letters, 6(4), pp. 251-64 Kumar D. and Ravi V. (2008). Predicting Credit Card Customer Churn in Banks Using Data Mining. International Journal of Data Analysis Techniques and Strategies, 1(1), pp. 4–28. Larivière B. and Van den Poel D. (2005). Predicting Customer Retention and Profitability by Using Random Forests and Regression Forests Techniques. Expert Systems with Applications, 29(2), pp. 472–84. Leeflang P.S.H., Tammo H.A.,van Doorn B.J., Hanssens D.M., van Heerde H.J., Verhoef P.C. and Wieringa J.E. (2009). Creating Lift versus Building The Base: Current Trends in Marketing Dynamics. International Journal of Research in Marketing, 26(1), pp. 13–20. Lessmann S., Baesens B., Mues C. and Pietsch S. (2008). Benchmarking Classification Models for Software Defect Prediction: A Proposed Framework and Novel Findings. IEEE Transactions on Software Engineering, 34(4), pp. 485–496. Levin N. and Zahavi J. (2001). Predictive Modeling Using Segmentation. Journal of Interactive Marketing, 15(2), pp. 2–22. Lima E., Mues C. and Baesens B. (2009). Domain Knowledge Integration in Data Mining Using Decision Tables: Case Studies in Churn Prediction. Journal of the Operational Research Society, 60, pp. 1096–1106. Lippmann R.P. (1987). An Introduction to Computing with Neural Nets. IEEE, ASSP Magazine, 4(2), pp. 4–22. Madden G., Savage S. and Coble-Neal G. (1999). Subscriber Churn in the Australian ISP Market. Information Economics and Policy. McCulloch W.S. and Pitts W.A. (1943). Logical Calculus of Ideas Immanent in Nervous Activity. Bulletin of Mathematical Biophysic 5, pp. 115–133. Mizerski R. (1982). An Attribution Explanation of the Disproportionate Influence of Unfavourable Information. Journal of Consumer Research, 9 (December), pp. 301– 310. Mozer M., Wolniewicz R., Grimes D., Johnson E. and Kaushansky H. (2000). Predicting Subscriber Dissatisfaction and Improving Retention in the Wireless Telecommunications Industry. IEEE Transactions on Neural Networks, 11(3), pp. 690–696. Neslin S.A., Gupta S., Kamakura W., Lu J. and Mason C.H. (2006). Measuring and Understanding the Predictive Accuracy of Customer Churn Models. Journal of Marketing Research, 43(2), pp. 204-211. Patuwo E.B., Hu M.Y. and Hung M.S. (1993). Two-Group Classification Using Neural Networks. Decision Science, 24(4), pp. 825–845.

5


Paulin M., Perrien J., Ferguson R., Salazar A. and Seruya L. (1998). Relational Norms and Client Retention: External Effectiveness of Commercial Banking in Canada And Mexico. International Journal of Bank Marketing, 16(1), pp. 24–31. Reinartz W.J. and Kumar V. (2003). The Impact of Customer Relationship Characteristics on Profitable Lifetime Duration. Journal of Marketing, 67, pp. 77-99. Risselada H., Verhoef P.C. and Bijmolt T.H.A. (2010). Staying Power of Churn Prediction Models. Journal of Interactive Marketing, 24, pp.198-208. Rumelhart D.E., Hinton G.E. and Williams R.J. (1986). Learning Representations by Back-Propagating Errors. Nature, 323, pp. 533– 536. Rust R.T. and Zahorik A.J. (1993). Customer Satisfaction, Customer Retention, and Market Share. Journal of Retailing, 69(2), pp. 193–215. Shaffer G. and Zhang J.Z. (2002). Competitive One-to- One Promotions. Management Science. 48(9), pp. 1143-60. Sharma A. and Panigrahi P.K. (2011). A Neural Network Based Approach for Predicting Customer Churn in Cellular Network Services. International Journal of Computer Applications, 27(11), pp. 26-31. Stum, D., & Thiry, A. (1991). Building customer loyalty. Training and Development Journal, 45(4), 34–36. Swets JA. (1988). Measuring the Accuracy of Diagnostic System. Science, 240, pp. 1285-1293. Tsai C.F. and Lui Y.H. (2009). Customer Churn Prediction by Hybrid Neural Network. Expert System with Applications, 36(10), pp. 2547-12553. Van den Poel D. (2003). Predicting Mail-Order Repeat Buying. Which Variables Matter? Review of Business and Economics, Katholieke Universiteit Leuven. Van den Poel D. and Larivière B. (2004). Customer Attrition Analysis for Financial Services Using Proportional Hazard Models. European Journal of Operational Research, 157(1), pp. 196–217. Verbeke W., Dejaeger K., Martens D., Hur J. And Baesens B. (2012). New Insights into Churn Prediction in the Telecommunication Sector: A Profit Driven Data Mining Approach. European Journal of Operational Research, 218, pp. 211-229. Verbeke W., Martens D., Mues C. and Baesens B. (2011). Building Comprehensible Customer Churn Prediction Models with Advanced Rule Induction Techniques. Expert Systems with Applications, 38, pp. 2354–2364 Wei C. and Chiu I. (2002). Turning Telecommunications Call Details to Churn Prediction: A Data Mining Approach. Expert Systems with Applications, 23, pp. 103– 112. Xie Y., Li X., Ngai E.W.T., and Ying W. (2008). Customer Churn Prediction Using Improved Balnaced Random Forests. Expert System with Applications, 36, pp. 5445449.

6


Yan L., Miller D.J., Mozer M.C. and Wolniewicz R. (2001). Improving Prediction of Customer Behavior in Nonstationary Environments, In Proceedings of the International Joint Conference on Neural Network, Washington, DC, 3, 2258-2263. Zahavi J. and Levin N. (1997). Applying Neural Computing to Target Marketing. Journal of Direct Marketing, 11(1), pp. 5–22. Zeithaml V., Berry L. and Parasuraman A. (1996). The Behavioural Consequences of Service Quality. Journal of Marketing, 60(2), pp. 31–46.

7