A Hidden Markov Model of Customer Relationship Dynamics

• Activity Number: 175
• Type: Invited
• Date/Time: Monday, August 8, 2005 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics and Marketing
• Author(s): Oded Netzer*+ and James Lattin and V. Srinivasan
• Companies: Columbia University and Stanford University and Stanford University
• Keywords: Customer relationship ; hidden Markov models ; dynamic choice models ; hierarchical Bayes ; direct marketing

Abstract

This research addresses the issue of modeling and understanding the dynamics of customer relationships. Our model facilitates using typical transaction data to evaluate the impact of customer-brand encounters on the dynamics of customer relationships and their subsequent buying behavior. In the proposed model, customer-brand encounters may have an enduring impact by shifting the customer to a different (unobservable) relationship state. We constructed and estimated a hidden Markov model (HMM) to relate the latent relationship states to the observed buying behavior. This model enables us to dynamically segment the firm's customers base and to examine methods by which the firm can use marketing actions to alter long-term buying behavior. We use a hierarchical Bayes approach to model and estimate the unobserved heterogeneity across customers. We calibrate the model in the context of alumni relations using longitudinal gift giving data provided by the Stanford Alumni Association. The application of the model for marketing decisions is illustrated using a "what-if" analysis of a reunion marketing campaign. Additionally, we demonstrate improved prediction ability on a validation sample.