Estimating Customer Segmentation based on Customer Lifetime Value Using Two-Stage Clustering Method

Abstract

In order to cope with the competitive environment related to beauty industry sector in Indonesia, companies need to manage and evaluate customer interactions by enhancing Customer Relationship Management (CRM). This study aims to specify customer segment that has similar lifetime value with clustering method, hence company can conduct appropriate strategies to the right segment. Two-stage clustering method for segmenting customers is proposed in this study. Ward's method is used for choosing an initial number of cluster and K-Means method to perform clustering analysis. Two approaches using LRFM (Length, Recency, Frequency, Monetary) model and extended model called LRFM - Average Item (AI) variables in clustering process are compared by validity index to obtain the best result for customer segmentation. The result shows that adding new variable Average Item in LRFM model have no significant difference or better results in clustering. The ranking process based on Customer Lifetime Value (CLV) score is conducted using weighted LRFM model variables. Final weight score for all variables are obtained from Fuzzy AHP method. In summary, company also get several inferences such as customer characteristics of high and less potential customers. It can be a guideline for making the sale and marketing strategies.