Understanding customer purchase behavior is essential for organizations seeking to strengthen customer engagement, refine marketing strategies, and improve business performance in highly competitive digital markets. This study develops a hybrid Customer Segmentation and Structural Equation Modeling (SEM) framework to examine the determinants of customer purchasing behavior and purchasing intention. The proposed framework integrates K-Means clustering with SEM, enabling the simultaneous identification of homogeneous customer groups and assessment of structural relationships among Customer Demographics (CD), Product Preference (PP), Purchasing Behavior (PB), Customer Segmentation (CS), and Purchasing Intention (PI). A dataset containing 8,000 customer records was analyzed through data preprocessing, exploratory data analysis, clustering, and structural modeling. The K-Means results identified four meaningful customer segments: Budget Customers, Regular Customers, Premium Customers, and High-Value Customers. The SEM results indicated a satisfactory overall model fit (CFI = 0.953, TLI = 0.947, RMSEA = 0.051, and SRMR = 0.043). Product Preference was found to exert a significant positive effect on Purchasing Behavior (β = 0.451, p < 0.001), while Purchasing Behavior demonstrated the strongest direct influence on Purchasing Intention (β = 0.536, p < 0.001). In addition, Customer Segmentation significantly mediated the relationship between Purchasing Behavior and Purchasing Intention. These findings confirm that combining machine-learning-based customer segmentation with SEM provides a robust analytical framework for understanding both customer heterogeneity and behavioral mechanisms. The proposed approach offers practical value for personalized marketing, customer targeting, resource allocation, and data-driven strategic decision-making across increasingly complex digital commerce environments.
