A comprehensive framework for superstore business with employing effective clustering techniques

Abstract

A superstore is an extensive store offering a diverse variety of everyday commodities under one roof, saving customers the trouble of shopping at different locations. The market industry is rapidly expanding, and to maximize profit utilizing customer behaviour, superstores need to constantly monitor their client's purchasing patterns and take appropriate measures to keep their loyalty while pushing them to spend more and bring in more new clients. The research presents a suitable framework for segmenting superstore consumers based on their attributes and assessing customer value through profit analysis applying appropriate segmentation and clustering techniques. An extensive comparison of the recency, frequency, monetary value (RFM) and length, recency, frequency, monetary value (LRFM) models employing three clustering algorithms: K-means Clustering, Agglomerative Clustering, and Fuzzy C-means Clustering is experimented to obtain the optimal framework. According to the findings, the LRFM model with the K-means algorithm produces the most promising output. It consists of 7 clusters where cluster 3 is the most crucial cluster as its customers hold the most customer value and generate the most profit for the superstore. As a result, superstore owners have a better understanding of their customers' needs and wants, allowing them to implement effective marketing strategies for the relevant sector.