Analisis Klasterisasi Pelanggan Layanan Pijat Menggunakan Algoritma K-Means Clustering Pada Griya Sehat Faza Depok

Abstract

Customer segmentation is a crucial element in strengthening data-driven marketing strategies, especially for Micro, Small, and Medium Enterprises (MSMEs) operating in the service sector. Customer data management at Griya Sehat Faza Depok is still conducted conventionally without adopting an analytical approach to identify consumer profiles. This limitation hinders the optimization of marketing strategies and efforts to maintain customer loyalty. This study groups home massage service customers using the K-Means Clustering algorithm. The CRISP-DM (Cross Industry Standard Process for Data Mining) framework, which includes business understanding, data understanding, data preparation, modeling, evaluation, and deployment, is used in this study. Customer transaction data from January to December 2025 serves as the dataset for analysis. Five variables are used in the clustering process: visit frequency, treatment type, total expenditure, service duration, and transportation costs representing customer distance. Data preprocessing stages include data cleaning, categorical data encoding, aggregation, and normalization using the Min-Max method. The optimal number of clusters is determined using the Elbow Method, while the quality of clustering results is evaluated using the Davies-Bouldin Index (DBI). The analysis results show the formation of four customer groups with a DBI value of 0.647, indicating good clustering quality. Each group exhibits distinct behavioral characteristics, enabling the identification of high-value customers, loyal customers, regular customers, and low-value customers. These findings offer practical recommendations for developing targeted marketing strategies, improving customer retention, and supporting more efficient therapist allocation at Griya Sehat Faza Depok.