Application of Market Basket Analysis to Redesign Supermarket Layouts for Optimum Returns in Kenyan Supermarkets

Abstract

Market Basket Analysis also known as Association Rule Mining is a data mining technique for discovering/ identifying interesting relations between variables in a large database. It helps businesses uncover relationships between frequently purchased products by consumers. This technique unveils insights into consumer purchasing behavior, facilitates the creation of compelling product bundles, and better shelve planning in a physical supermarket. Market Basket Analysis aims to identify patterns and associations among products by analyzing transactional data, such as customer purchase history or shopping cart contents. This study explores the application of Market Basket Analysis (MBA) to optimize supermarket layouts for enhanced profitability in Kenyan supermarkets. Utilizing both the Apriori and ECLAT algorithms, the research aims to identify the most frequently purchased items, explore items commonly bought together, and compare the performance of these two models in computing item relationships. The research adopts a descriptive design to analyze transactional data from the supermarket. Transactional data from a Kenyan supermarket database was extracted and preprocessed for analysis. Utilizing a minimum support threshold of 0.01 and a minimum confidence level of 0.6, 223 frequent item sets and 96 rules were generated. This analysis identified several significant association rules, such as {HARPIC, SUGAR WHITE} => {VIM POWDER} with a support of 0.0106 and confidence of 0.7079, and {SUGAR WHITE, VIM POWDER} => {HARPIC} with a support of 0.0106 and confidence of 0.7895. Additionally, the rule {JAMAA B/S, PASTA TAK, SUGAR WHITE} => {AJAB NGANO} was identified with the support of 0.0112 and confidence of 0.8827, among others. This meant that these items can be placed on the shelves next or close to each other as they are highy co-related. This will enhance their visibility increasing there chances of being bought by customers thus increasing sales as well meeting the customers varying preferences. The insights gained from this analysis ensures better product placement on the shelves.The findings provide actionable insights for supermarket managers to enhance cross-selling opportunities, optimize store layouts, and improve profitability and customer satisfaction.