Fractal Dimension of Shopping Path: Influence on Purchase Behavior in a Supermarket

This paper aims to use fractal dimensions to quantify the complexity of customer in-store movements, and proposes a purchase model factoring in the effects of complex customer movements on purchase behavior. We used the box-counting method to calculate the fractal dimension of shopping paths and investigated its relationships with basket size and sales, which are viewed as important for marketing. We found that the customer group with high fractal dimensions had mean values for the number of sales floor zones visited, basket size, stay time in store, and sales amount statistically higher than those of the customer group with lower fractal dimensions. We analyzed a binomial logit model to identify positive effects that the fractal dimension has on purchases in the food ingredient categories of vegetable, fish, and meat. (C) 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license