Multi-criteria suitability analysis and spatial interaction modeling of retail store locations in Ontario, Canada

Abstract

GIS-based decision analysis is increasingly used by retailers to address the complexity and cost of investment in retail store location decisions. This study conceptualizes and represents nine criteria in a GIS-based multi-criteria decision analysis of 4.7 million potential retail store locations. From topographic statistics to spatial interaction modelling, the study utilizes criteria of varied complexity to analyze the statistical and spatial distribution of highly suitable locations for a retail store. The study further examines how the spatial representations of criteria based on the Huff model affects the distribution of suitable locations. The results show that although Toronto dominates the retail landscape in Ontario, key regions are found in Guelph, Kitchener-Waterloo and Cambridge. Results show that the incorporation of network-based spatial interaction costs in Huff’s model produces more spatially heterogeneous sales estimates than Euclidean-based spatial interactions. Future research efforts in improving various components of the suitability analysis, as well as the scaling and regional parameterization of spatial interaction models are also discussed.