Pemasaran Efektif melalui Optimasi RFM Analisis Berdasarkan K-Means Clustering

Abstract. The digitalization trend demands an accurate understanding of customer behavior. RFM (Recency, Frequency, Monetary) Analysis with K-Means Clustering offers objective segmentation for big data, overcoming the limitations of traditional scoring methods. This study aims to identify customer purchasing patterns of PT Maha Nagari Nusantara and formulate effective marketing strategies. This quantitative-exploratory study analyzed 86,605 customer transactions of PT Maha Nagari Nusantara on Shopee (2024). The data was processed into RFM, then Min-Max normalization was performed. Exploration of k values with values ​​of 2 to 6 with evaluation of results using the Silhouette Coefficient, choosing k = 2 (0.5906) as the most objective data structure. K-Means identified two main segments, namely "Active & Potential Customers" as many as 40,404 customers with low Recency an average of 100 days with Frequency-Monetary values ​​above average. Prioritize strategies with loyalty building, upselling, cross-selling. “Passive & Risky Customers” totaling 46,201 customers with a very high Recency of 270 days on average indicates that they have been inactive for a long time, at risk of churn. Strategy priority with reactivation campaigns and increased engagement. A differentiated dual marketing strategy, focusing on active retention and passive reactivation, is essential company to optimize budgets and increase retention. Abstrak. Tren digitalisasi menuntut pemahaman perilaku pelanggan yang akurat. RFM (Recency, Frequency, Monetary) Analisis dengan K-Means Clustering menawarkan segmentasi objektif untuk data besar, mengatasi keterbatasan metode skoring tradisional. Penelitian ini bertujuan mengidentifikasi pola pembelian pelanggan PT Maha Nagari Nusantara dan merumuskan strategi pemasaran efektif. Penelitian kuantitatif-eksploratif ini menganalisis 86.605 transaksi pelanggan PT Maha Nagari Nusantara di Shopee (2024). Data diolah menjadi RFM, lalu dilakukan normalisasi Min-Max. Eksplorasi nilai k dengan nilai 2 hingga 6 dengan evaluasi hasil menggunakan Silhouette Coefficient, memilih k=2 (0,5906) sebagai struktur data paling objektif. K-Means mengidentifikasi dua segmen utama yakni “Pelanggan Aktif & Berpotensi” sebanyak 40.404 pelanggan dengan Recency rendah rata-rata 100 hari dengan nilai Frequency-Monetary di atas rata-rata. Prioritas strategi dengan pembinaan loyalitas, upselling, cross-selling. “Pelanggan Pasif & Berisiko" berjumlah 46.201 pelanggan dengan Recency sangat tinggi rata-rata 270 hari menandakan sudah lama tidak aktif, berisiko churn. Prioritas strategi dengan kampanye reaktivasi dan peningkatan keterlibatan. Strategi pemasaran ganda yang terdiferensiasi, berfokus pada retensi aktif dan reaktivasi pasif, esensial bagi perusahaan untuk optimasi anggaran dan peningkatan retensi.