Cluster analysis in marketing

Methods of cluster analysis are statistical methods, which are being studied and applied across a wide range of fields, from biology and economics, to astrology and marketing. Hence, it is crucial for such methods to be described properly in professional literature as well as in study materials, in order for it to be used adequately. This thesis is a reaction to the book Statistics for Marketing and Customer Research, by Mario Mozzaachi, where the author presents statistical methods used in marketing. Mario Mozzaachi claims that the choice of distance does not affect the outcome of hierarchical cluster analysis. The goal of this thesis to provide an overview of statistical methods used in the field of marketing; and secondly to examine the effect of distance choice in the cluster analysis process. The first goal is achieved by providing an overview of ten statistical methods used in marketing. By describing their fundamentals and application in practice. Whereas, the second goal is full filled by demonstrating the importance of distance selection in clustering process on an illustrative example, by creating a marketing persona. For this purpose three different linkage methods in combination with seven distance/similarity measures were applied in programming language R. Clustering solutions are then evaluated and compared, while contradiction of Mario Mozzaach’s claim is demonstrated.