<![CDATA[Sales increase is an essential factor for telecommunication service providers, including ICONNET, a subsidiary of PLN, amid intense market competition. Companies face the challenge of designing effective marketing strategies without structured customer data analysis. This study aims to apply the K-Means Machine Learning algorithm to analyze and cluster ICONNET customer data in Bangka Belitung, with the expected results supporting strategic sales increase decisions. The methodology employed is Data Mining with the CRISP-DM framework, where the modeling process implements the K-Means algorithm. The determination of the optimal number of clusters (K) was consistently performed using the Elbow Method and Silhouette Score, yielding an optimal value of K=2. The clustering results successfully divided customers into two main groups: Cluster 0, dominated by users of low-value packages (Package 1 and 2), and Cluster 1, consisting of users of higher-value packages (specifically Package 5). This segmentation provides a basis for ICONNET to formulate differentiated service strategies and targeted marketing offers tailored to the characteristics and preferences of each customer segment, which directly supports operational efficiency and long-term business growth.]]>
Copyrights © 2025
