<![CDATA[This study aims to formulate marketing strategies in the printing industry through customer segmentation based on Recency, Frequency, and Monetary (RFM). The research problem is the absence of transaction behavior-based customer mapping that can distinguish loyal customers, high-value customers with declining activity, new or potential customers, low-active customers, and churned/passive customers. The novelty of this study lies in integrating active and churned customer separation, RFM indicator construction, comparison of K-Means and K-Medoids methods, and translation of cluster results into operational STP and CRM strategies in the printing industry context. This study applies a descriptive quantitative approach involving data cleaning, order-level deduplication, customer anonymization, RFM construction, active and churned customer separation, data transformation and normalization, clustering, model evaluation, and marketing strategy interpretation. The results show that of 11,268 customers, only 529 customers (4.69%) are active, while 10,739 customers (95.31%) are categorized as churned or passive. The best model is K-Means with k=4, DBI 0.8366, and Silhouette Score 0.4364. The segmentation produces four active customer groups used as the basis for loyalty, high-value customer retention, repeat purchase improvement, and customer reactivation strategies. ABSTRAKPenelitian ini bertujuan merumuskan strategi pemasaran pada industri percetakan melalui segmentasi pelanggan berbasis Recency, Frequency, dan Monetary (RFM). Permasalahan penelitian adalah belum adanya pemetaan pelanggan berbasis perilaku transaksi yang mampu membedakan pelanggan loyal, pelanggan bernilai tinggi yang mulai menurun aktivitasnya, pelanggan baru atau potensial, pelanggan aktif rendah, dan pelanggan churned/pasif. Kebaruan penelitian ini terletak pada integrasi pemisahan pelanggan aktif dan churned, pembentukan indikator RFM, pembandingan metode K-Means dan K-Medoids, serta penerjemahan hasil cluster ke dalam strategi STP dan CRM yang operasional pada konteks industri percetakan. Penelitian menggunakan pendekatan kuantitatif deskriptif dengan tahapan pembersihan data, deduplikasi order, anonimisasi pelanggan, pembentukan RFM, pemisahan pelanggan aktif dan churned, transformasi dan normalisasi data, clustering, evaluasi model, serta interpretasi strategi pemasaran. Hasil menunjukkan bahwa dari 11.268 pelanggan, hanya 529 pelanggan (4,69%) yang aktif, sedangkan 10.739 pelanggan (95,31%) tergolong churned atau pasif. Model terbaik adalah K-Means dengan k=4, DBI 0,8366, dan Silhouette Score 0,4364. Segmentasi menghasilkan empat kelompok pelanggan aktif yang menjadi dasar strategi loyalitas, retensi pelanggan bernilai tinggi, peningkatan pembelian ulang, dan reaktivasi pelanggan.]]>
