<![CDATA[Customer segmentation is an essential strategy for understanding customer behavior and supporting effective customer relationship management. However, many customer segmentation studies focus primarily on grouping customers based on behavioral characteristics without considering the economic value of each customer segment. This study proposes a customer segmentation framework that integrates the Length, Recency, Frequency, and Monetary (LRFM) model, Self-Organizing Maps (SOM), and Customer Lifetime Value (CLV) analysis to provide a more comprehensive understanding of customer behavior and value. The study utilized customer transaction data from PT XYZ consisting of 11,233 transaction records involving 7,937 customers during the 2024 observation period. Customer transaction data were transformed into LRFM attributes, normalized using Min-Max Normalization, and clustered using the SOM algorithm. The clustering quality was evaluated using Silhouette Score and Davies-Bouldin Index (DBI), while CLV analysis was employed to assess the economic contribution of each customer segment. The results indicate that the optimal SOM configuration successfully generated three customer segments with a Silhouette Score of 0.6026 and a DBI value of 0.5644. Based on CLV analysis, the identified segments were classified as Core Customers, New Customers, and Lost Customers. The findings demonstrate that each segment exhibits distinct behavioral and economic characteristics, enabling more targeted customer management strategies. The integration of LRFM, SOM, and CLV provides a comprehensive customer segmentation framework that supports customer prioritization, customer relationship management, and data-driven marketing decision-making.]]>
Copyrights © 2026
