Modellierungsansätze in der Warenkorbanalyse im Überblick

By now, most retailers have launched efforts to commit customers by means of customer card programs and the accompanying benefits. As a result, retailers have access to large amounts of customer data, because those customer cards allow them to generate a complete purchase history for every single customer. Answering strategic questions, such as those concerning customer relationships, should bring about a sustainable advancement of each customer relationship. In this retailing application, market basket analysis is generally used. Market basket analysis is a method of investigating the composition of a consumer's shopping basket on a single shopping occasion. In this context, selection decisions in multiple categories are assumed to be related. In a modeling framework, considering these cross-category linkages is a prerequisite for deriving suitable marketing strategies. This article reviews relevant publications on market basket analysis and classifies them into exploratory and explanatory approaches, focusing more on the explanatory approaches. Complementarity, heterogeneity and coincidence are identified as factors which determine multi-category choices and explain intercategory relationships. © Wirtschaftsuniversität Wien, Austria 2006.