Improving Customer Relationship Management Models by Including Neighborhood Effects Using Spatial Econometrics

• Activity Number: 224
• Type: Topic Contributed
• Date/Time: Monday, August 1, 2011 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics and Marketing
• Author(s): Philippe Baecke*+ and Dirk Van den Poel
• Companies: Ghent University and Ghent University
• Keywords: predictive analytics ; customer acquisition ; CRM ; spatial models ; neighborhood effects ; spatial correlation

Abstract

Customer relationship management (CRM) uses data mining techniques to support decisions about several marketing strategies such as customer acquisition, development and retention. From these three applications, customer acquisition models suffer the most from the fact that only a limited amount of information is available about potential prospects resulting in a relative low accuracy of the models. This study tries to improve the predictive performance of such models by including neighborhood effects. Traditional CRM models ignore the fact that correlation, resulting from social influences and homophily, could exist between the purchasing behaviors of surrounding prospects. Hence, a spatial model is used to capture these neighborhood effects and improve the identification of prospects for automobile brands. Moreover, an optimization is presented concerning the measurement level of the neighborhoods.