<![CDATA[The telecommunications industry faces challenges in understanding customer characteristics due to large data volumes and diverse service usage behaviors. Customer segmentation becomes a strategic approach to support more targeted and effective marketing strategies. This study aims to apply the K-Means clustering algorithm to perform data-driven customer segmentation in a telecommunications company. The customer dataset undergoes preprocessing stages including missing value handling, categorical variable encoding using One-Hot Encoding, and feature scaling with StandardScaler. The optimal number of clusters is determined using the Elbow Method. The results show the formation of three customer segments with distinct characteristics based on tenure, monthly charges, total charges, and churn rate. Visualization using Principal Component Analysis (PCA) clearly illustrates the separation among clusters. An interesting finding reveals that the segment with the highest customer value also has the highest churn rate, indicating the need for more specific retention strategies. The contribution of this study lies in providing an unsupervised learning-based customer segmentation approach that can assist companies in designing more effective marketing strategies, improving customer retention, and supporting data-driven decision making.]]>
Copyrights © 2025
