OPTIMIZING RETAIL STRATEGY WITH APRIORI ALGORITHM FOR INFORMED DECISION-MAKING ON CUSTOMER PURCHASING PATTERNS

Abstract

In retail business, understanding customer purchase patterns is crucial for enhancing marketing strategies and product placement. This research focuses on analyzing product sales transactions at Minang Mart Lubuk Begalung branch, aiming to identify frequently purchased product combinations and understand the relationships between products through association rules. Data, subjected to preprocessing stages including data cleaning, outlier handling, and normalization, ensures consistent quality. Frequent itemset analysis using the Apriori algorithm with varying minimum support thresholds (0.003, 0.005, and 0.008) provides insights into customer purchase patterns. Association rules with a minimum confidence level of 0.5 and a minimum lift level of 1 yield significant findings, such as the combination of ADES NATURAL 24X600ML and CHEETOS JAGUNG BAKAR 40X40 GR. From these findings, Minang Mart can design more effective marketing strategies and enhance product placement in sales areas. Visualization of purchase patterns through graphs supports a more intuitive understanding of customer preferences. The research results are expected to contribute positively to optimizing sales strategies and strengthening the competitiveness of retail businesses in this era of intense competition.