A Comparative Study of RFM-Based Clustering Methods in Customer Segmentation

To exist in the present competitive market, companies should focus on their customers, understand their needs, and fulfill them. Providing value to the customer is a necessary thing for a company. They should know well about their customer’s characteristics. For understanding the characteristics of customers, segmenting them according to their previous sales, history, demographics, family, and relationships has become very popular in this context. Data mining approaches are now widely adopted by many organizations, which is helping them generate sales and sustain themselves in a competitive market. In this paper, an analysis is done of over 23053 transactions of 5506 customers and segments the customers according to their recency, frequency, and monetary (RFM) values. 3 different clustering algorithms have been applied and compared for the segmentation, and the final model outperformed the previous studies. The proposed algorithm will help the companies to understand their customer’s behaviors and segment them into different categories, which will lead them to set up proper marketing strategies for their consumers.