Combination of RFM’s (Recency Frequency Monetary) Method and Agglomerative Ward’s Method for Donors Segmentation

Customer segmentation is a technique to improve the company's relationship with its customers. The RFM method is often used when segmenting customers. The RFM (Recency Frequency and Monetary) method is used to determine the character of each customer based on the transaction data made. Each variable is then given a weight to determine the final score. Final scores are used to divide customers into several segments. This research will replace the scoring procedure with RFM-based cluster analysis. Ward's Method is a Hierarchical Agglomerative Clustering technique that has high memory and time complexity. The cluster technique has the advantage of maximizing the homogeneity of group members and minimizing variance within the group. Customers will be segmented using this clustering algorithm and RFM. The data used is 10070 data and is divided into 2 datasets. Data is partitioned to reduce memory usage and outlier handling. Outlier data has 991 data and normally distributed data has 9079 data. With a maximum evaluation result of 0.936 in the distribution of 2 clusters for outlier data and a maximum evaluation result of 0.345 in the distribution of 3 clusters for normal distribution data. It follows that outlier data has significant heterogeneity between groups, while data with a normal distribution has low heterogeneity.