K-Means Implementation for Customer Segmentation in Aluminium and Glass Business by Purchasing Behavior

Abstract

— Mulia Jasa Aluminium dan Kaca is a business in the retail and service sector, offering Aluminium and glass materials and services for manufacturing, installation, and repair. Currently, competition in this field is quite intense, leading the business owner to admit difficulties in increasing sales. Therefore, the business owner needs to implement marketing and service strategies to boost sales. However, the diversity of customers with varying characteristics and behaviors makes it challenging to establish effective marketing and service strategies. Thus, this study conducts customer segmentation based on purchasing behavior. The aim is to understand customer behavior and loyalty using sales report data from the business. The variables used to assess a customer's value are Length, Recency, Frequency, and Monetary (LRFM). These variables are grouped using the K-means clustering algorithm. The objective of this study is to group customers based on their purchasing behavior, thereby assisting the business in developing more effective marketing and service strategies, enhancing customer satisfaction, and ultimately increasing sales and loyalty. Using the Silhouette method to determine the optimal number of clusters, three customer groups were identified, with the highest coefficient value of 0.663063. Cluster 0 is the “Lost Customer Group”, Cluster 1 is the “New Customer Group”, and Cluster 2 is the “Core Customer Group”.