The development of digital technology has brought significant changes to people’s lifestyles, including in shopping habits, which are now largely conducted online. One of the sectors directly affected is the online food business, where entrepreneurs need to implement the right strategies to remain competitive and increase sales. This research was conducted on Dapur Tiara, a food business that markets its products through Instagram and WhatsApp. So far, transaction data has only been stored as an archive, even though it has great potential to be analyzed in order to uncover consumer purchasing patterns.
This study applies the Apriori algorithm to analyze sales transaction data and identify products that are frequently purchased together. The main objective is to generate association rules that can serve as the foundation for cross-selling strategies. The process includes data cleaning, transformation into basket format, generation of frequent itemsets, evaluation of association rules using FP-Growth and ECLAT as comparisons, and presenting the results through a web application built with Flask. Experiments were conducted with minimum support values ranging from 0 to 1 with an increment of 0.1, and minimum confidence values ranging from 0 to 1 as well.
The results show that several product combinations have strong relationships, such as Mini Pizza and Pepperoni Pizza with a confidence of 0.48 and a lift of 1.872. The developed web application allows users to upload transaction data and automatically display the analysis results in the form of tables and recommendation sentences. This system can be used to design data-driven sales strategies, increase the transaction value per customer, and provide more personalized and relevant product recommendations.
Keywords: Apriori Algorithm, Sales Transactions, Dapur Tiara, Product Recommendation, Association Analysis, Web Visualization