This study uses the K-Nearest Neighbors (K-NN) method to predict and classify the best-selling products in a hardware store. With the current development of information technology, sales trend analysis and prediction have become an important part of the business decision-making process. The popular K-NN classification algorithm is used to analyze sales data from a public dataset to determine which products are most in demand by consumers. The process of data collection, selection, preprocessing, transformation, data mining, and evaluation of results are all part of the Knowledge Discovery in Database (KDD) stages. The analysis results show that products in the “active” category sell more than products in the “passive” category. Out of the total data, 56 were successfully categorized as active data, and the remaining 29 were categorized as passive data. This study is expected to provide deeper insights into consumer behavior and assist building material store management in making better decisions using the data they possess. This is anticipated to enhance the company's competitiveness and improve operational efficiency.