Sales Data Classterization Analysis Using K-Means Method for Marketing Strategy Development

In the digital era, utilizing sales data is very important to support strategic decision making. This research aims to overcome the problems faced by 9Doors Store in optimizing marketing strategies and stock management. The main problem faced is the lack of in-depth analysis of existing sales data, which results in difficulties in formulating appropriate marketing strategies and efficient stock management. For this reason, this research applies the K-Means Clustering method to group products based on customer purchasing behavior characteristics. The data used includes product categories, selling prices, initial stock, number of products sold, and total sales obtained from 9Doors Store during the period March to September 2024. The method used in this research is Data Mining approach with K-Means algorithm, which is implemented using RapidMiner software. The data analysis process goes through Knowledge Discovery in Databases (KDD) stages, including data collection, data cleaning (preprocessing), data transformation, and data mining using K-Means. Cluster evaluation is done using Davies-Bouldin Index (DBI) to assess the quality of clustering results. The results of this study show that the division of sales data into three clusters provides optimal results with the lowest DBI value (0.106), which indicates efficient clustering. This finding identifies products with high, medium, and low sales levels, which can be used to formulate more targeted marketing strategies. With these results, Toko 9Doors can improve stock management and design more effective promotions based on better customer segmentation.