<![CDATA[The rapid expansion of the e-commerce industry in Indonesia has led to a substantial increase in consumer behavioral data. This condition highlights the need for analytical approaches capable of identifying transaction patterns and supporting meaningful customer segmentation. This study aims to segment e-commerce customers using the K-Means clustering method and to present the results through an interactive dashboard. The research process includes data cleaning and standardization, followed by the construction of derived variables such as average basket, spend per month, and paylater share to represent customer transaction behavior. The optimal number of clusters was determined by evaluating several metrics, including the Sum of Squared Errors (SSE) graph, Silhouette Score, Calinski–Harabasz Index, and Davies–Bouldin Index. Although the custest Silhouette Score and Calinski–Harabasz Index were obtained at k = 2, the overall evaluation indicated that k = 3 produced a more balanced and interpretable clustering structure. Consequently, three customer segments were identified: customers with moderate shopping activity, high-value customers with a tendency to use Paylater services, and customers with high transaction frequency but relatively low purchase value. The analytical results were subsequently implemented in a Streamlit-based interactive dashboard to support data-driven decision making.]]>
Copyrights © 2025
