Supermarket, a shop that provides various products for use, especially for daily life, including food products, drinks, kitchen utensils, clothing, electronic equipment and others. It is not surprising that many mothers now choose to shop for daily necessities at supermarkets rather than the nearest stall. With self-service, it can make it easier for us consumers to buy different products in one place. So there is no need to change shops to buy other items. Of course, products have different levels of popularity, not only because of taste but also because of price. The number of products provided by supermarkets is relatively large and if you look at the level of popularity, it is difficult to determine, so data mining is needed. The data mining used is clustering. After implementing and using the K-Means algorithm in clustering (grouping) supermarket products, there are two centroids used (C1 for Not Selling Products and C2 for Best Selling Products). The initial centroid value is determined randomly, while the subsequent centroids are adjusted according to the results of calculating the closest distance (maximum distance). The final result obtained is that the best-selling group consists of 12 products, namely products with serial numbers 1, 4, 5, 6, 7, 8, 9, 11, 14, 15, 16 and 17. Meanwhile, the product group does not There are 6 best-selling products, namely products with serial numbers 2, 3, 10, 12, 13 and 18.