The increasing availability of supermarket scanner data permits much richer studies of consumer demand for supermarket products. Most of the product categories consist of a large number of differentiated goods. Using a comprehensive supermarket data set, I explore several alternative methodologies for modeling consumer demand in the presence of differentiated products. I provide three applications in which consumer demand provides a basis for understanding and developing optimal firm strategy and competition policy.
In the first chapter, I develop and apply the popular discrete choice model to char acterize consumer segmentation and to develop retailer strategy that exploits the knowledge of these segments. This model applies to any product category in which consumers systematically purchase a single unit of a single product alternative on any given trip. In the second chapter, I present an alternative model of demand that allows for multiple unit shopping. I develop the importance of accounting for multiple-item purchases in terms of the predictions for substitution patterns and managerial strategy Finally, in the third chapter, I use the multiple-unit purchase model to investigate the consequences of mergers in the carbonated soft drink industry. In this final section, I compare my results to those obtained from using the discrete choice model.
My Research