Uncovering Customer Archetypes in Direct-to-Consumer Apparel: A K-Means Clustering Analysis of DMart Sales Data

In the evolving direct-to-consumer (DTC) apparel sector, understanding customer behavior beyond traditional demographic segmentation is essential for enabling personalized marketing and optimizing inventory decisions. This study applies K-means clustering to analyze historical sales data from DMart, focusing on uncovering actionable customer segments based on transactional behavior. The dataset includes purchase frequency, monetary value, product preferences, return rates, and channel usage. A hybrid analytical approach integrating Exploratory Factor Analysis (EFA), Confirmatory Factor Analysis (CFA), Structural Equation Modeling (SEM), and K-means clustering was employed. The findings reveal four distinct customer archetypes: Practical Loyalists, Seasonal Sporadic, Value-Driven Explorers, and Comfort-First Seniors. The study demonstrates how machine learning combined with multivariate analysis enhances segmentation accuracy and supports targeted marketing strategies.