Consumer Segmentation With K-Means at Lucky Shop Tanjungbalai

Abstract

Consumer segmentation is an important strategy for improving marketing effectiveness and inventory management in retail businesses. Lucky Shop Tanjungbalai faces challenges in understanding diverse customer purchasing patterns, making it difficult to develop targeted marketing strategies. This study aims to apply the K-Means Clustering method to classify consumers based on purchasing behavior patterns. The data used consisted of 15 customer transaction records collected from Lucky Shop Tanjungbalai, with attributes including purchase frequency, quantity of purchased products, and product categories. This research adopted a qualitative approach combined with data mining techniques using the CRISP-DM framework, which consists of business understanding, data understanding, data preparation, modeling, evaluation, and deployment. The system was developed using PHP and MySQL. The results indicate that K-Means Clustering successfully segmented customers into Loyal Customers and Occasional Customers based on their purchasing characteristics. These segmentation results provide practical benefits for Lucky Shop by enabling more targeted promotional programs, improving customer relationship strategies, optimizing inventory planning, and supporting data-driven business decision-making. Therefore, the implementation of K-Means Clustering can serve as an effective solution for customer segmentation in local retail businesses.

Keywords

Data Mining; K-Means Clustering; Consumer Segmentation; Marketing Strategy; PHP; MySQL