Defection detection: Measuring and understanding the predictive accuracy of customer churn models

被引:304
作者
Neslin, SA [1 ]
Gupta, S
Kamakura, W
Lu, JX
Mason, CH
机构
[1] Dartmouth Coll, Amos Tuck Sch Business Adm, Hanover, NH 03755 USA
[2] Columbia Univ, Grad Sch Business, New York, NY 10027 USA
[3] Duke Univ, Fuqua Sch Business, Durham, NC 27706 USA
[4] Comerica Banck, Detroit, MI 48226 USA
[5] Univ N Carolina, Kenan Flagler Business Sch, Chapel Hill, NC 27515 USA
关键词
D O I
10.1509/jmkr.43.2.204
中图分类号
F [经济];
学科分类号
02 ;
摘要
This article provides a descriptive analysis of how methodological factors contribute to the accuracy of customer churn predictive models. The study is based on a tournament in which both academics and practitioners downloaded data from a publicly available Web site, estimated a model, and made predictions on two validation databases. The results suggest several important findings. First, methods do matter. The differences observed in predictive accuracy across submissions could change the profitability of a churn management campaign by hundreds of thousands of dollars. Second, models have staying power. They suffer very little decrease in performance if they are used to predict churn for a database compiled three months after the calibration data. Third, researchers use a variety of modeling "approaches," characterized by variables such as estimation technique, variable selection procedure, number of variables included, and time allocated to steps in the model-building process. The authors find important differences in performance among these approaches and discuss implications for both researchers and practitioners.
引用
收藏
页码:204 / 211
页数:8
相关论文
共 29 条
[1]  
ALKER HAYWARD, 1965, MATH POLITICS
[2]  
[Anonymous], MARK LETT
[3]  
[Anonymous], MARKETING SCI
[4]  
Berry M.J., 2000, MASTERING DATA MININ
[5]   Dimension reduction in binary response regression [J].
Cook, RD ;
Lee, H .
JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 1999, 94 (448) :1187-1200
[6]  
FARLEY JU, 1986, GEN MARKET RESPONSE
[7]  
FITCHARD K, 2002, TELEPHONY ONLIN 0318
[8]  
FRIEDMAN J, 2003, RECENT ADV PREDICTIV
[9]   Greedy function approximation: A gradient boosting machine [J].
Friedman, JH .
ANNALS OF STATISTICS, 2001, 29 (05) :1189-1232
[10]   MULTIVARIATE ADAPTIVE REGRESSION SPLINES [J].
FRIEDMAN, JH .
ANNALS OF STATISTICS, 1991, 19 (01) :1-67