Segmenting and targeting customers through clusters selection & analysis

Abstract

This paper investigates the use of machine learning clustering technique to segment and target customers of a wholesale distributor. It describes the selection, analysis, and interpretation of clusters for evaluating customers annual spending on the products. We show how circular statistics can categorize customers by looking at the annual spending on six essential product categories. Several clusters were created using k-means clustering algorithm and an in-depth analysis on these clusters were performed using several techniques to carefully select the best cluster. Automated clustering was able to suggest groups that these customers fall into. The evaluation and interpretation of clusters were able to provide insights into various purchase behaviors and to nominate the best customer group to target.