Using Shopping Baskets to Cluster Supermarket Shoppers

Today, competition forces consumer goods manufacturers and retailers to differentiate from their competitors by specializing and by offering goods/services that are tailored towards one or more subgroups or segments of the market. Retailers understand that shoppers are heterogeneous in nature, that they possess different wants and needs and that it is impossible to satisfy them all. However, the retailer in the FMCG sector is highly limited in his ability to segment the market and to focus on the most promising segments, since the typical attraction area of the retail store is too small to afford neglecting a subgroup within the store’s attraction area. In fact, the supermarket should appeal to as many of the heterogeneous public in its attraction area as possible. Nevertheless, given the huge amount of transactional and loyalty card data being collected, this leads to the intriguing question whether in-store segmentation may be a viable alternative to discover homogeneous customer segments based on their shopping behaviour. Indeed, the discovery of homogeneous customer groups who tend to use the store similarly could be used for instance to target individuals with customized promotions. Therefore, in this paper we will introduce a new methodology for behaviour-based customer segmentation. More precisely, we will use the method of latent class mixture modelling to discover hidden customer segments on the basis of the contents of their shopping baskets. Furthermore, loyalty card data is used to find out if these customer segments differ in terms of socio-demographic or lifestyle characteristics and whether these characteristics can be used to target different customer segments with more relevant product offers. The method is carried out on market basket data from a major Belgian supermarket store. Results indicate that a number of distinct segments can be identified who differ significantly in terms of the average purchase rate within a pre-determined set of product categories.