Executive Summary
Multipurpose shopping is a prominent and relevant feature of shopping behavior. To empirically assess how the demand for multipurpose shopping depends on retail agglomeration, this article proposes and illustrates a model of the combined choice of purpose and destination of shopping trips. The authors estimate the model on data obtained from a large regional shopping survey in the Netherlands that focused on shopping center choice for three types of goods: groceries, clothing and shoes, and other semidurable goods. For each good, the authors model the choice of shopping trip purpose and shopping destination. The modeling results demonstrate how agglomeration variables affect shopping behavior and how sales of one good are affected by the other goods present in the shopping center.
Analyses such as those reported in this article can help assess if and how retailers can benefit from being located in the proximity of other types of stores. Parameters in the model represent different types of agglomeration effects. First, the authors estimate purpose adjustment effects, or the extent to which shopping-trip purpose choices vary with retail supply. Second, the authors measure the between-store attraction effects, or the extent to which attributes of stores that offer other goods than the focal good contribute to the shopping center’s utility. The authors found that the location and size of a shopping center have a significant impact on the choice of trip purpose. Furthermore, it appears that the perceived purpose-specific utilities of shopping destinations exhibit significant and positive joint-attraction effects. An unexpected finding in the study was that attraction effects are also produced by stores that were irrelevant to the purpose of the shopping trip.
This suggests that different store types all contribute to destination attractiveness, even if no purchases are made from these stores. To assess the extent to which these forces affect shopping center choice probabilities, the authors compared the prediction accuracy among multipurpose and single-purpose models for each focal good. It appeared that the single-purpose models systematically overpredict the market share of the smaller centers and underpredict the market share of the larger centers for two of the three good categories.
Our findings suggest that groceries and other goods benefit more from the presence of clothing and shoes stores than vice versa. Such effects are obscured in single-purpose models. Retail managers should therefore consider using multipurpose models such as the one developed in this article instead of single-purpose models to predict the impact of changes in retail supply on shopping trips. There is sufficient scope for this given that the retail-supply and shopping trip data required to estimate the proposed multipurpose model are fairly straightforward to collect.
Finally, the proposed model structure provides a framework for modeling consumer choice in a variety of situations in which agglomeration effects occur. The authors demonstrate the approach with an application to shopping center choice, but the approach can also be used to model individual store choice or even item choice in the context of basket selection.
Biography
Theo A. Arentze received an MSc degree in cognitive psychology from Groningen University and a Ph.D. in urban planning from Eindhoven University of Technology. He is now Associate Professor in the Urban Planning group at the same university. His main fields of expertise and current research interests are decision support systems, behavioral modeling, knowledge discovery, and learning with applications in urban planning and marketing.
Harmen Oppewal is Professor of Retail Marketing, Department of Marketing, Monash University, Melbourne, Australia. He held academic appointments at the University of Surrey (UK), University of Sydney (Australia), and Eindhoven University of Technology in The Netherlands. He received his PhD from Eindhoven University of Technology and holds degrees in Geography (BSc) and Psychology (BSc, MSc) from the University of Groningen. His research focuses on understanding and modeling consumer decision-making behaviour in various application domains, in particular retailing.
Harry J.P. Timmermans received a Ph.D. in spatial sciences from the Catholic University of Nijmegen. He is Chair of the Urban Planning Group; Director of the European Institute of Retailing and Consumer Services; and Dean of the Faculty of Architecture, Building, and Planning at the Eindhoven University of Technology. His main fields of expertise concern behavioral modeling, consumer studies, and computer systems in a variety of application contexts, including transportation.
J Marketing Research, Volume 42, Number 1, February 2005
View Table of Contents.