Predictive Modeling for Consumer Purchase Decisions

• Activity Number: 7
• Type: Invited
• Date/Time: Sunday, August 8, 2004 : 2:00 PM to 3:50 PM
• Sponsor: ASA, Southern Ontario Chapter (Toronto)
• Author(s): Zhen Mei*+
• Companies: Manifold Data Mining, Inc.
• Keywords: predictive modeling ; cluster analysis ; consumer behavior

Abstract

Consumer purchase decision is the process of selecting, purchasing, using, or disposing of products, services, ideas, or experiences to satisfy the consumers needs and desires. As the economy has been shifting from product-driven to customer-focused, understanding consumer behaviors has become an integrated part of business and marketing operations for many companies. Consumer purchase behavior is influenced by many factors, such as culture, social class, reference groups and family, and psychological influences, and so on. Demographics, lifestyle, and household expenditures are the key quantifiable variables for describing and understanding consumer behaviors. We will introduce a hybrid algorithm of cluster analysis and parametric modeling to predict the likelihood of consumers in responding favorably to a given product. By leveraging the predictive power of Manifold comprehensive micro-marketing databases and an adaptive dimension reduction technique we are able to link consumer behavior with the demographic data: age, family, education, dwelling, occupation, employment, income, ethnicity, etc., and household expenditure data on food, clothing, shelter, recreation, etc.