This dissertation studies the potential of analytical CRM for firms operating in non-contractual settings. Specifically, it addresses the questions of how to expand customers’ relationship breadth, i.e., customers’ cross-buying behavior, and how to predict customers’ purchase activity and intensity, i.e., customers’ relationship length and depth. With respect to relationship breadth, the results reveal that cross-buying across tangible product categories is driven fundamentally different than cross-buying services. With respect to predicting relationship length and depth, empirical results reveal that current state-of-the-art stochastic models fail to outperform simple management heuristics. However, latest developments in the field of machine learning, specifically support vector machines, solve these problems efficiently. On the basis of theses results, this dissertation gives implications for research and practice.
«This dissertation studies the potential of analytical CRM for firms operating in non-contractual settings. Specifically, it addresses the questions of how to expand customers’ relationship breadth, i.e., customers’ cross-buying behavior, and how to predict customers’ purchase activity and intensity, i.e., customers’ relationship length and depth. With respect to relationship breadth, the results reveal that cross-buying across tangible product categories is driven fundamentally different than cr...
»