<![CDATA[In today’s highly competitive retail landscape, businesses face increasing challenges in retaining customer loyalty and achieving sustainable growth. A common issue, particularly among small and medium-sized enterprises (SMEs), is the absence of a structured method for identifying and categorizing customers based on their value and behavior. This study addresses the challenge by implementing a data-driven customer segmentation approach using Recency, Frequency, and Monetary (RFM) analysis combined with the K-Means clustering algorithm. The research utilized 2,353 transaction records from 369 unique customers collected over three years from a local retail business. After preprocessing and normalizing the RFM values using Min-Max scaling, the Elbow Method was applied to determine the optimal number of clusters, resulting in four distinct customer segments. Cluster 3, labeled “Loyal Customers,” consisted of customers with high purchase frequency and very high spending; Cluster 1 (“Potential Loyalists”) included those with moderate activity; Cluster 0 represented “At-Risk Customers,” and Cluster 2 comprised “One-Time Buyers.” This segmentation framework supports the development of targeted Customer Relationship Management (CRM) strategies, such as loyalty programs and re-engagement campaigns. However, the approach also has limitations, including potential data bias due to the use of static transaction records and the challenge of interpreting clusters without qualitative customer feedback. Despite these constraints, the study demonstrates the practical utility of combining RFM analysis with clustering techniques to extract actionable insights in environments with limited technical infrastructure.]]>
Copyrights © 2025
