Empirical Investigation of Retail Expansion and Cannibalization in a Dynamic Environment

M of retail chains who seek to add new stores or close existing ones need to know the net impact of a store’s opening/closure on the overall chain performance. This requires inferring the extent to which each store generates incremental sales as opposed to competing with other stores belonging to the chain for the same set of customers. However, when the chain is experiencing a growth or a decline in sales, not accounting for these dynamics in goodwill is likely to yield misleading estimates of incremental sales versus cannibalization. Moreover, firms might have been strategic in opening outlets in locations with favorable characteristics. We need to control for this location endogeneity while inferring the marginal effect of store opening/closure. In this paper, we develop a demand model that accounts for dynamics in goodwill, location endogeneity, and spatial competition between geographically proximate retail outlets. We calibrate the model parameters on both attitudinal and behavioral data for a fast food chain in a large U.S. city. The results imply that consumers perceive a travel cost of $0.60 per mile. As regards the composition of sales at individual stores, on average, 86.7% of sales constitute incremental purchases with the rest derived from cannibalized sales from nearby stores belonging to the chain. We also find significant decay in cannibalization with distance such that when the distance between stores increases by one mile, the sales lost due to cannibalization decreases by 28.1%; there is virtually no cannibalization at a distance of 10 miles. In terms of managerial applications, we discuss how managers can use the model presented in this paper to make two key decisions: (a) isolating locations that can be closed by identifying stores that yield the lowest marginal benefit to the chain and (b) dealing with franchisees’ potential concerns about cannibalization.