COMPARATIVE ANALYSIS OF K-MEANS AND K-MEDOIDS ALGORITHMS IN DETERMINING CUSTOMER SEGMENTATION USING RFM MODEL

A poultry farm company has customers across various regions and needs customer segmentation for a more effective marketing strategy. This research aims to perform the company customer segmentation through a comparative analysis using K-Means and K-Medoids algorithms. Customer segmentation was performed using the Cross-Industry Standard Process for Data Mining (CRISP -DM) technique on three distinct sales datasets. We selected the attributes of the datasets based on the Recency-Frequency-Monetary (RFM) model. The number of clusters was established utilizing the Elbow approach, and the assessment of clustering was carried out using the Davies-Bouldin Index (DBI) methodology. Through a comparative examination, it becomes evident that the superiority of the K-Means algorithm is demonstrated in comparison to the K-Medoids algorithm. This outcome is derived from the observation that the DBI value for K-Means is lower than the DBI value for K-Medoids across the three sales datasets. Based on these results, it is recommended to determine customer segmentation using the K -Means algorithm. We obtained four customer segments: superstar, typical, newcomer, and dormant. The superstar segment has the most recent recency value, the highest frequency, and the most significant monetary value. The results of this customer segmentation became a consideration for the company in making marketing strategies that are right on target. This study can be used as a reference for current issue in the poultry farm company for supporting and preparing food, supports current issue in the sustainable development goals (SDGs).