Statistics and data mining techniques for lifetime value modeling
D.R. Mani, James Howard Drew, Andrew L. Betz et al. · ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD) · 1999
Customer lifetime value (LTV)-which is a measure of the profit generating potential, or value, of a customer-is increasingly being considered a touchstone for customer relationship management.The central challenge in predicting LTV is the production of estimated customer tenures with a given service supplier, based on information contained in company databases.Classical survival analysis techniques like proportional hazards regression focus on covariate effects frequently presumed to be linear, and examination of age-wise effects can be difficult.Further, segments of customers, whose lifetimes and covariate effects can vary widely, are not necessarily easy to detect.A new neural network model for hazard prediction is used to free proportional hazards-like models from their linearity and proportionality constraints, and clustering tools are applied to identify segments of customer hazard patterns.Using the proportional hazards and neural network models in tandem, we demonstrate how data mining tools can be apt complements of the classical statistical models, and show that their combined usage overcomes many of the shortcomings of each separate tool setresulting in a LTV tenure prediction model that is both accurate and understandable.
Cited by 61 works. doi:10.1145/312129.312205 Open-access PDF
Authors
References
- Nonparametric Estimation from Incomplete Observations
- Nonparametric Estimation from Incomplete Observations
- Regression Models and Life-Tables
- Regression Models and Life-Tables
- Neural Networks: A Comprehensive Foundation
- Multilayer feedforward networks are universal approximators
- Multilayer feedforward networks are universal approximators
- Analysis of Survival Data
- Analysis of Survival Data.
- Statistical Models and Methods for Lifetime Data