Customer lifetime value determination based on RFM model

Abstract

Purpose – One of the salient challenges in customer-oriented organizations is to recognize, segment and rank customers. Customer segmentation is usually based on customer lifetime value (CLV) measured by three purchase variables: “Recency,” “Frequency” and “Monetary.” However, due to the ambiguity of these variables, using deterministic approach is not appropriate. For tackling this matter, the purpose of this paper is to propose a new method of customer segmentation and ranking by combining fuzzy clustering (as a segmentation method) and fuzzy AHP (as a ranking method). Design/methodology/approach – First, customers are classified based on purchase variables using fuzzy c-means clustering algorithm. Second, the variables are weighed applying an optimized version of AHP method. Considering the derived weights and customer groups, this paper follows to ranks segments based on CLV. The developed methodology has been implemented for a large IT company in Iran. Findings – The results show a tremendous capability to the company to evaluate his customers by dividing them into nine ranked segments. The validity of clusters has been submitted. Research limitations/implications – For researchers, this study provides a useful literature by combining FCM and an optimized version of fuzzy AHP in order to cover the limitations of previous methodologies. For organizations, this study clarifies the procedure of customer segmentation by which they can improve their marketing activities. Practical implications – Managers can consider the proposed CLV calculation methodology for selling the next best services/products to the group of customers that are more valuable, by calculating the entire lifetime value of the customers. Originality/value – This study contributes to the process of customer segmentation based on CLV, proposing a new method which covers the limitations of previous customer segmentation methods.