<![CDATA[The development of transportation and application-based services highlights the importance of user behavior analysis as a basis for data-driven marketing strategies. This study analyzes the segmentation of GOJEK service users (GoRide, GoCar, and GoFood) among students of the Faculty of Engineering, Sultan Ageng Tirtayasa University (FT UNTIRTA) using the Recency, Frequency, and Monetary (RFM) approach. Primary data were collected through questionnaires distributed to 105 active GOJEK users using purposive sampling. Data were analyzed through pre-processing, standardization, determination of the optimal number of clusters using the Elbow method, and clustering using the K-Medoids algorithm, which was selected over K-Means and K-Median due to its robustness against outliers, suitability for non-normally distributed RFM data, and use of actual data points as cluster centers for more interpretable segmentation results. The results showed that the optimal number of clusters for each service was three, classified as loyal, active, and passive customers. In GoRide, the distribution was 15 loyal, 32 active, and 16 passive users; in GoCar, 16 loyal, 10 active, and 35 passive users; and in GoFood, 25 loyal, 1 active, and 52 passive users. Loyal clusters are characterized by low recency and high frequency and monetary values, active clusters show medium usage rates, and passive clusters exhibit low frequency and transaction values. These results demonstrate that the RFM and K-Medoids combination is effective in identifying behavioral differences among GOJEK users, as validated by the Silhouette Score and Davies-Bouldin Index confirming cluster compactness and separation quality, and can serve as a basis for formulating more targeted marketing strategies in the student environment.]]>
