One of the consequences of the intense transformation experienced by the retail sector in recent decades has been the diversification of store formats. Although definitions of these store formats, or types, are often inexact and sometimes confusing, they reveal the increasing variety of store models in the ever more heterogeneous retail market. Also, while academic research has focused more on customer choices within store format, evaluating competition across store formats is an understudied area. Store formats determine a retail competitive structure. They define the pattern by which stores may be grouped based on the degree of overlap among their target segments (i.e., basis of the intensity of the competitive interaction). This structuring entails distinguishing between an inter-format competitive level that relates to the rivalry between store formats, and an intra-format competitive level that relates to the rivalry among stores of the same format. Therefore, our goal in this paper is to analyze the role of retail store format in spatial competitive interaction. In this study, we incorporate the possible asymmetric effects of different store formats. In this respect, we extend traditional models of spatial analysis and capture the spatial competitive interactions of store formats.

We use data from Salamanca, Spain, where we collected information about shopping behavior by surveying households. We asked respondents for a list of stores at which he or she usually shopped and a subjective estimation of the allocation of expenditure to each store during the previous month. For the explanatory variables considered in our model, we obtained measurements at the store level from the Alimarket Supermarket Census and the MOSAIC Iberia geographic information system. The former provided the location and size of each store, as well as a classification into supermarkets, hypermarkets, and discount stores. The latter provided the means to obtain a spatial interpretation of the study scenario by assigning geographic coordinates to consumers and stores. Consistent with the aggregate focus of this study, the location of consumers and stores can be distinguished at the census-track level. The results support our hypotheses, i.e., (i) the rivalry within store formats is significantly different from rivalry across store formats and (ii) that there is more rivalry within store formats relative to the rivalry across store formats. This implies a two-step hierarchy in the process of retail store choice in which the consumer chooses first the type of store in which to shop and second, the specific store within that format. The spatial dimension is significant in the second level of decision making.

Our results are relevant to retailers who want to create a portfolio of store formats and a network of stores for which market coverage is optimized and cannibalization effects are minimized. To obtain this optimization, those retailers must anticipate the consequences of new locations, closures, or competitors’ diversification strategies. In this regard, this study demonstrates that spatial competitive analysis should not be isolated from the effect of other attraction dimensions because the effect of location depends on the retail positioning of competitors, specifically, on the degree the target segments overlap. In addition, this study shows that spatial analysis should not focus exclusively on a specific store format. Although intra-format competitive interaction is more intense, the consequences of inter-format rivalry should not be dismissed.