Customer Segmentation Using RFM Analysis and Unsupervised Machine Learning: A Data-Driven Framework for Strategic Marketing

In the highly competitive e-commerce landscape, personalized marketing strategies have become vital for customer retention and profitability. This study presents a comprehensive, data-driven customer segmentation framework combining the classical RFM (Recency, Frequency, Monetary) model with unsupervised machine learning algorithms. The proposed method applies logarithmic transformation to mitigate data skewness and employs K-Means, DBSCAN, and Agglomerative Clustering to identify optimal customer clusters. Using the Elbow and Silhouette methods, the model determines the most suitable segmentation approach. The resulting clusters-High-Value Loyalists and Low-Engagement Customers-enable strategic marketing actions for enhanced customer engagement. This work highlights how integrating traditional marketing analytics with modern machine learning techniques can extract actionable business intelligence from transactional data.