Inventory is an important component in a company's operational activities, especially in the trade sector, because it directly affects the smoothness of sales and the level of customer satisfaction. Unstructured inventory management can lead to stockpiling of goods, stock shortages, and inappropriate decision making. Maisa Building Materials Store located in Semerap, Kerinci Regency, Jambi, currently still records inventory manually, so the shop owner has difficulty in identifying items with high and low sales levels. This study aims to implement the K-Means Clustering algorithm in grouping inventory based on sales levels to support more effective and efficient stock management. The research method used is data mining with the stages of data collection, preprocessing, manual calculation of the K-Means algorithm, and implementation using RapidMiner software. The analyzed data amounted to 507 inventory items that have gone through a data cleaning process so that they are suitable for use in grouping. Grouping is done with two clusters, namely a cluster of goods with a low sales level and a cluster of goods with a high sales level. The results of the study indicate that 494 items, or 97.44 percent, fall into the low-sales cluster, while 13 items, or 2.56 percent, fall into the high-sales cluster. These results indicate that most products have relatively low sales turnover, while only a small proportion contribute significantly to total store sales. The information generated from this clustering process can be used as a basis for decision-making in inventory management, particularly in determining stocking priorities, stock control, and developing appropriate, data-driven marketing strategies.