Application of K-Means for Product Grouping Best Sellers at Planet Tire Jatibarang Branch

Abstract

This research aims to identify the best-selling products at Planet Tire Workshop Jatibarang Branch using the K-Means Clustering method. Understanding product sales patterns is important in designing effective marketing strategies and managing stock efficiently. This research uses sales transaction data for one year, including the number of sales, product types, and total transaction value. The analysis process includes data preprocessing, selection of relevant attributes, application of the K-Means algorithm, and validation of the optimal number of clusters with the Elbow method. As a result, products were grouped into three categories: high, medium, and low sales. The high sales cluster contributes significantly to revenue, while the medium sales cluster shows potential for improvement through promotion, and the low sales cluster requires further evaluation. This research helps management manage stock, prioritize promotions, and optimize resource allocation. However, the research has limitations as it has not considered external factors such as seasonal trends and promotions, and focuses on one branch. Development of the research in other branches can expand its benefits. The results of this study are expected to improve operational efficiency, support data-driven strategies, and enrich academic literature related to the application of K-Means in retail management and sales data analysis.