Pertumbuhan minat investasi seperti perhiasan emas di Indonesia memberikan peluang yang baik bagi para pemilik usaha perhiasan emas, termasuk Usaha Menengah, Kecil, dan Mikro (UMKM). Seiring dengan hal tersebut, perkembangan zaman dan teknologi mengarahkan perubahan pola belanja masyarakat Indonesia menuju pembelanjaan online melalui e-commerce. Dengan meningkatnya daya saing, UMKM perlu membangun customer relationship management guna meningkatkan loyalitas pelanggan. Untuk itu, penelitian ini bertujuan untuk mengetahui perilaku pelanggan melalui segmentasi pelanggan dan merancang strategi pengembangan pelanggan UMKM Perhiasan Emas. Metode K-Means Clustering digunakan untuk mengelompokkan pelanggan menjadi beberapa klaster berdasarkan variabel Recency, Frequency, dan Monetary (RFM). Metode Customer Lifetime Value (CLV) dan Customer Value Matrix (CVM) juga digunakan untuk mengetahui karakteristik pelanggan pada setiap klaster. Setiap klaster akan dirancang strategi pengembangan dengan penentuan prioritas menggunakan Complex Proportional Assessment (COPRAS). Terdapat empat klaster pelanggan yang terbentuk dalam penelitian ini dengan delapan rekomendasi strategi peningkatan loyalitas pelanggan berdasarkan karakteristik masing-masing klaster yang terbentuk.

---

The growing interest in investments such as gold jewelry in Indonesia provides good opportunities for gold jewelry business owners, including Medium, Small and Micro Enterprises(MSMEs). Along with this, the times and technology have directed changes in spending behavior of the Indonesian people towards online shopping through e-commerce. With increasing competitiveness, MSMEs need to develop customer relationship management to maintain and increase customer loyalty. For this reason, this study aims to determine customer behavior through customer segmentation and design customer development strategies for Indonesian Gold Jewerly MSMEs. The K-Means Clustering method is used to group customers into several clusters based on Recency, Frequency, and Monetary variables. Customer Lifetime Value(CLV) and Customer Value Matrix(CVM) methods are also used to determine the characteristics of customers in each cluster. Each cluster will be designed with a development strategy with prioritization using the Complex Proportional Assessment(COPRAS). In this study, there are four clusters with eight recommendation strategies to increase customer loyalty based on the characteristics of each cluster.