Statistical Evaluation of the Doughnut Clustering Method for Product Affinity Segmentation

Product affinity segmentation is a powerful technique for marketers and sales professionals to gain a good understanding of customers’ needs, preferences, and purchase behavior. Performing product affinity segmentation is quite challenging in practice because product level data usually have high skewness, high kurtosis, and large percentage of zero values. The Doughnut clustering method has been shown to be effective using real data, and was presented at SAS Global Forum 2013 (Baer & Chakraborty, 2013). However, the Doughnut clustering method is not a panacea for addressing the product affinity segmentation problem. There is a clear need for a comprehensive evaluation of this method in order to be able to develop generic guidelines for practitioners on when to apply the method. In this paper, we meet the need by evaluating the Doughnut clustering method on simulated data with different levels of skewness, kurtosis, and percentage of zero values. We developed a five-step approach based on Fleishman’s power method to generate synthetic data with prescribed parameters. Subsequently, we designed and conducted a set of experiments to apply the Doughnut clustering method as well as the traditional K-means method as benchmark on the simulated data. We draw conclusions on the performance of the Doughnut clustering method by comparing the clustering validity metric “the ratio of between-cluster variance to within-cluster variance” as well as the relative proportion of cluster sizes against those of K-means. In certain data situations, the Doughnut clustering method is shown to produce an acceptable clustering solution when other approaches fail.