Spatial Shopping Behavior in a Multi-Channel Environment: A Discrete Choice Model Approach
Thomas Wieland
DOI: https://doi.org/10.18335/region.v8i2.361
2021-08-04
REGION
Abstract:Spatial impacts of online shopping are discussed frequently in retail geography. Here, online shopping is mostly regarded as a central driver of competition for physical retailing and its locations, such as town centers or malls. Due to its high popularity, cross-channel shopping is sometimes considered to be a support for physical retailing. However, traditional retail location theory does not consider shopping channels other than in-store shopping. Furthermore, although online shopping is far too important to be neglected in examining consumer spatial shopping behavior, there is an obvious lack in the previous literature towards incorporating multi- and cross-channel shopping into store choice models. The present study aims to identify the main drivers of store choice on the basis that both in-store and online shopping alternatives are available, as well as the opportunity for cross-channel shopping. Taking into account previous literature on both physical store choice and multi-channel shopping, hypotheses on the impact of different shopping transaction costs (such as travel time, delivery charges, or uncertainty with respect to the stores' assortment) were derived. Based on a representative consumer survey, real past shopping decisions in three retail sectors (groceries, consumer electronics [CE], and furniture) were collected. The econometric analysis of empirical store choices was performed using a nested logit model which includes both physical and online stores. The results confirm several assumptions of classical retail location theory as well as previous findings from single-firm studies and stated choice experiments on multi-channel shopping behavior. Travel time to physical stores reduces consumer utility and store choice probability, respectively. Consumer sensitivity towards travel time decreases with decreasing purchase frequency of the desired goods. Delivery charges also decrease the likelihood of choosing a store. The impact of cross-channel integration on store choice (assuming the reduction of consumer transaction costs) is considerably lower than expected and differs between retail sectors. While furniture retailers profit from enabling cross-channel shopping, there is no such competitive advantage found for grocery and CE retailers. The positive effect of assortment on condition of diminishing marginal utility is confirmed for grocery stores and CE stores, but not for furniture stores. From a theoretical perspective, this study shows that multi- and cross-channel shopping behavior does not contradict the main thoughts of classical retail location theory. From a practical perspective, the study is a contribution as store choice models play a significant role in both business location planning and governmental land use planning.