Optimization of Customer Segmentation with RFM, K-Means, and FP-Growth for Marketing Strategy

Snap Digital Printing has experienced a 6% decline in sales since 2019 and a continuous decline in the number of customers until 2023. The main cause is the company's limited use of customer data, which hinders the identification of behavioral changes, weakens market response, and reduces loyalty. This study aims to evaluate the impact of inadequate customer data management on sales and customer numbers, while analyzing the effectiveness of the RFM model, K-Means algorithm, and FP-Growth in customer segmentation, purchase pattern analysis, and marketing optimization. Data was collected from the Ownshop Snap Digital Printing branch, covering the period from January 2019 to December 2023, with 7,203,059 records before cleaning and 7,029,561 after cleaning. The analysis identified three customer segments: Low Engagement 86%, Active 3%, and VIP 11%. The FP-Growth results showed average Support of 11.36%, 6.73%, and 7.85%; Confidence of 85.70%, 49.72%, and 82.60%; and Lift Ratios of 1.77, 7.67, and 4.92. The findings indicate that data-driven systems strengthen customer-focused marketing strategies, enhance retention and sales, and provide a solid foundation for evidence-based practices in the digital printing industry.