Implementasi Data Mining dengan Metode K – Means Clustering Guna Menentukan Strategi Pemasaran Produk Konveksi

Abstract

The Indonesian garment industry is growing rapidly, but Zaya Konveksi in Gresik faces significant sales decline. Despite the advancement of technology and e-commerce, Zaya Konveksi has not fully utilized these potentials to promote its products and directly interact with consumers. This study aims to develop a data-driven customer segmentation model to enhance marketing effectiveness for SMEs in the garment industry. A quantitative approach using K-means Clustering was applied to categorize respondents into three Clusters based on attributes like age, location, and shopping preferences. Data was processed using RapidMiner, with evaluations conducted through centroid distance calculations, normalization, and iterative centroid updates to ensure internally homogeneous and externally heterogeneous Clusters. Marketing decisions based on Clusters reveal distinct characteristics and preferences. Cluster 0 focuses on women aged 22–27 in Gresik, promoting black t-shirts priced at IDR 150,000–300,000 over 16 days via TikTok, emphasizing price and quality. Cluster 1 targets women aged 16–27 in Sidoarjo, promoting yellow hoodies over 14 days using Instagram Ads, prioritizing appealing designs. Cluster 2 targets students aged 16–21 in Sidoarjo, promoting blue hoodies and raglans over 15 days via Instagram, emphasizing brand and product aesthetics. Social media-based Clustering strategies significantly improve targeted marketing and customer loyalty.