Analisis Pola Pembelian Pelanggan dengan Algoritma Frequent Pattern Growth (Studi Kasus: Cafe Kopi Boss Jember)

Abstract

Customer purchasing pattern analysis plays a crucial role in developing business strategies. Based on observations, Cafe Kopi Boss Jember has not yet utilized historical sales data to understand customer purchasing behavior, even though such data can support business decision-making. This study aims to implement the FP-Growth algorithm to analyze purchasing patterns at Cafe Kopi Boss Jember and generate promotional package recommendations. This applied research uses association rule mining with the FP-Growth algorithm, which constructs an FP-Tree to discover patterns among menu items. The minimum support is determined using the adaptive support method with an adjustment coefficient α, resulting in a minimum support of 0.01 and a minimum confidence of 20%. The implementation resulted in two association rules that met the criteria. The first rule, “milkshake matcha → french fries,” has a support of 0.017, confidence of 0.213 and a lift ratio of 1.401. The second rule, “kubisu → joshua,” has a support of 0.016, confidence of 0.311 and a lift ratio of 1.735. These results indicate that customers who purchase milkshake matcha tend to buy french fries, while those who purchase kubisu are likely to buy joshua. Based on these patterns, two bundling packages are proposed: a light snack package (milkshake matcha and french fries) and a favorite drinks package (kubisu and joshua). These recommendations are expected to enhance the marketing strategy of Cafe Kopi Boss