Maximizing Profits for a Multi-Category Catalog Retailer

It is a common trend in the retail industry for catalog retailers to mail multiple catalogs, each promoting different product categories. The existing catalog mailing models do not address the issue of optimizing multi-category catalog mailing. We address this research gap by introducing a model that integrates the when and what components of a customer's purchase decision into the how much component (number of catalogs) of a firm's cross-selling strategy. In addition to comparing the impact of category-specific versus full product catalogs in generating sales in a specific category, the study also finds relative impacts of various category-specific catalogs. We jointly estimate the probability of purchase and purchase amounts in multiple product categories by using multivariate proportional hazard model (MVPHM) and a regression based purchase amount model in a Hierarchical Bayesian framework. The model accounts for unobserved heterogeneity, and uses a control function (CF) approach to account for endogeneity in catalog mailing. The results from the Genetic Algorithm (GA) based optimization suggest that the catalog mailing policy as per the proposed model would be able to generate 38.4% more customer lifetime value (CLV) from a sample of 10% of the households as compared to the current catalog mailing policy of the retailer by reallocation of the catalogs across customers and mailing periods based on their propensity to buy. [PUBLICATION ABSTRACT]

Journal of Retailing
George, Morris; Kumar, V; Grewal, Dhruv
Document Type: 

Leave a comment

Comment Policy
This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.