Analisis Performa Penjualan Elektronik Menggunakan Metode RFM dan Neural Network (Studi Kasus PT. Makmur Sumber Redjeki)

Abstract

The development of information technology and increasing business competition require companies to effectively analyze historical sales data to understand customer behavior and improve marketing strategies. This study aims to analyze the electronic sales performance of PT. Makmur Sumber Redjeki by combining the Recency, Frequency, Monetary (RFM) method and a Neural Network model. The RFM method is used to group customers based on the last transaction time, purchase frequency, and transaction value, while the Neural Network model is used to more accurately classify customers into loyal, potential, and at-risk churn categories. The research data consists of customer sales transactions over a one-year period, processed using a supervised learning approach. The analysis stages include data normalization, customer segmentation using RFM, and training a Neural Network model using RFM features as input. The results of the study show that the integration of the RFM method with Risk 38.5%, Loyal 30.8% and Potential 30.8% and Neural Network is able to increase the accuracy of 75% in identifying strategic customer groups and producing useful insights in supporting data-based marketing decision making.