Implementation of the FP-Growth Algorithm for Bundling Strategy and Store Layout Redesign at Toko Kasih Ibu

Abstract

Grocery stores like Toko Kasih Ibu face increasing challenges in staying competitive against modern markets offering better convenience, product variety, and services. A notable sales decline in 2024 highlights the need for improved marketing and store layout strategies. This study analyzes purchasing patterns using the FP-Growth algorithm within a Market Basket Analysis (MBA) framework to design product bundling and optimal layout recommendations. Using the CRISP-DM approach, 468,507 transaction records from 2022–2024 were processed, followed by data preparation and transformation. The FP-Growth model was applied with a minimum support of 2% and confidence of 50%, resulting in 11 strong association rules—such as bundling Fom Burger Per 10, Pilus SP 500 RTG BAL, and Indomie Goreng PC. Additionally, category-level analysis using the Activity Relationship Chart (ARC) with the AEIOUX scale suggested reorganizing the store into four sectors to improve customer convenience and encourage combined purchases. The findings demonstrate that applying the FP-Growth algorithm with appropriate parameters offers valuable insights for effective bundling and layout strategies, supporting promotional efforts and sales goals.