UD. Penyubur Tani is a trading business that sells various kinds of needs for farmers in running their business in agriculture such as fertilizers, pesticides, seeds, and others. UD. Penyubur Tani wants to increase customer loyalty by giving rewards in the form of discounts so that its business is increasingly trusted and increasing, but in giving rewards to customers is still not effective, because UD. Penyubur Tani has difficulty in calculating one by one which customers whose frequency of purchases and total purchase price are most in the category of very loyal and loyal. Therefore, an application is needed to classify customers so that in determining strategies in building loyalty on target. By using the concept of CRM and the K-Means method, the results obtained from data processing are able to group customers who must be prioritized and can determine which customers deserve a reward. From 100 customer data, the K-Means Clustering method succeeded in grouping very loyal criteria by 8%, loyal 34% and potential by 58%.