<![CDATA[Improper drug inventory management can cause stockouts, overstocking, andinefficient procurement decisions, especially in small-scale retail pharmaciesthat still rely on manual estimation. This study analyzes drug sales patternsand groups drug inventory at Anugrah Pharmacy Bekasi using the K-Meansclustering algorithm within the Cross Industry Standard Process for DataMining (CRISP-DM) framework. The dataset consisted of 5,312 sales transactionsfrom January to June 2025. The transaction records were aggregatedinto 731 drug items using three variables: transaction frequency, sales volume,and transaction value. Data preparation included aggregation, missing-valuechecking, duplicate checking, transformation, and Min-Max normalization.The optimal number of clusters was determined using the Elbow Method,which indicated three clusters (k = 3). The K-Means results grouped the 731drug items into 28 Fast Moving items (3.83%), 129 Medium Moving items(17.65%), and 574 Slow Moving items (78.52%). The centroid analysis showsthat each cluster has distinct sales-movement characteristics. The results cansupport inventory decision-making by helping the pharmacy prioritize replenishmentfor fast-moving drugs, maintain controlled stock for medium-movingdrugs, and limit excessive procurement for slow-moving drugs. This studydemonstrates that CRISP-DM and K-Means clustering can provide practicalinformation for data-driven drug inventory management in a retail pharmacycontext.]]>
