A Model Development to Predict Customer's Defection in a Bank

다국어 초록 (Multilingual Abstract)

This paper has developed a model to predict customers’ defection and to apply the result of the model into a bank's marketing and IT strategies. This paper utilized data mining techniques to develop a model that can predict the defection rate of a bank's customers. Data mining is defined as a technique to extract useful information from large amounts of data and to find out relationships, patterns, and rules among data in order to use them in various corporate decisions makings. A logit, one of artificial neural network tools, and a decision treemodel are used to predict customers’ defection in this paper. This study selected more variables that are related to customers’ account information of a bank as significant predictors that influence on the customers’ defection.
The result of this study revealed meaningful facts that other banks would take them as lessons. That is, the more customers who have cancelled their deferred savings accounts after the account matured, and the greater the number of the deferred savings accounts customers currently have, the lower the tendency of defection of customers from a bank. The result also indicated that customers who had higher rate of average balance in deferred savings accounts when it’s compared to aggregate amount of the deposits, or customers who had higher number of demand deposits tend to defect from a bank more easily. The model provides banks with marketing tools such as cross-selling, calculation of customer's profit contribution, evaluation of customer's lifetime value, and customer segmentation to manage their customers. The result of the model also gives a meaningful lesson to IT field, too. The tools used to predict customers’ defection in this paper were useful, but analytical tool users have to be very careful to utilize various variables to improve the model's prediction power.

참고문헌 (Reference)

1 이호영, "시장세분화 모형 결합을 활용한 효과적인 생명보험 고객이탈예측모형에 관한 연구" 한국보험학회 74 (74): 33-58, 2006

2 Jones, T. O, "Why Satisfied Customer Defect" 88-99, 1995

3 Kang, H., "Seonggeon L" Jayoo Publish 2010

4 Kotler, P, "Principles of Marketing" Prentice-Hall 1996

5 Rumelhart, D. E., "Learning Internal Representations by Err Propagation, in Parallel Distributed Processing: Exploration in the Microstructure of Cognition" MIT Press 1986

6 Shaw, M. J., "Knowledge Management and Data Mining for Marketing" 31 : 127-137, 2001

7 Hirose,M, "Good bookmaking Pres" Kotler Marketing 2010

8 Reichheld, "F.. F. Loyalty Rules" Harvard Business School Press 2003

9 Ben-Akiva, M, "Discrete Choice Analysis: Theory and Application to Travel Demand" The MIT Press 1990

10 Han, Jiawei, "Data Mining: Concepts and Techniques" Morgan Kaufmann Publisher 2001

11 Wiesel, T., "Customer Equity: An Integral Part of Financial Reporting" 72 : 1-14, 2008

12 김상용, "CRM 고객데이터 분석을 통한 이탈고객 연구" 한국마케팅학회 7 (7): 21-42, 2005

13 Quinlan, J. R, "C 4.5: Programs for Machine Learning" Morgan Kaufman Publisher 1997

14 Lee, K. C., "An Artificial Intelligence-based Data Mining Approach to Extracting Strategies for Reducing the Churning] date in Credit Card Industry" 8 : 15-35, 2002

15 Fornell, C., "A National Customer Satisfaction Barometer: The Swedish Experiences" 55 : 1-, 1992