The Examining Consumer Behavior Patterns with Distance-Based Clustering Models in Business Analytics

Business analytics uncovers complex consumer behaviors by extracting structured models from customer data. Applying K-means and hierarchical clustering to a retail dataset containing demographic and spending variables, this study identifies latent customer segments. Although the Elbow method favored fewer clusters, the Gap Statistic and Silhouette index established an optimal five-cluster structure. To validate this configuration, oneway ANOVA (η2) demonstrated that cluster membership explains over 67% of the segmentation variance while Principal Component Analysis (PCA) confirmed its structural stability by retaining 78% of the total variance. Empirical results indicate that a five cluster solution provided the most meaningful partition, revealing distinct consumer clusters such as high-income conservative spenders and young groups of customers with elevated spending tendencies. Methodologically, the convergence of K-means and hierarchical clustering cross-validated the structural robustness of the segments, supported by high bootstrap stability (ARI = 0.817). Ultimately, this research bridges data-driven consumer insights with distance-based segmentation, highlighting the role of cluster analytics in enhancing business performance.