Customer Relationship Management for Supermarkets Using KMeans and RFM

Customer Relationship Management (CRM) is a process in which a business manages its interactions with customers using data analysis regarding the customers’ behaviors. CRM is essential for businesses across industries since it helps improve financial performance. Current studies have improved customer segmentation by either introducing a more comprehensive theoretical model or increasing the layers of machine learning models. Although studies have proved the need for supermarkets to incorporate CRM on a theoretical level, there is a lack of quantitative research on CRM for supermarkets. This paper proposes to combine K-Means with the RFM model to increase the explicability of the clusters and the CRM strategy. The dataset used in this paper contains consumer information for a supermarket. First, the K-Means model is used to cluster the customers with similar demographic backgrounds and purchase behaviors. Then, we include the RFM model to classify the different groups based on their overall value to the supermarket. As a result, we recommend selectively using promotions, product recommendations, e-mails, and specific campaigns to target and communicate with different groups.