Association Rules Mining for Designing Product Bundling Strategy (Study Case at a Cafe in Jakarta)

Abstract

Cafe X is a modern and prominent coffee shop in town that sells specialty beverage products with a total of 73 variations. Sales transactions that occurred in March 2021 at that Cafe reached an average of 300 times per day. All this time, the transaction data is only used to calculate inventory replenishment and to observe sales profits whereas it can actually be utilized to reveal information related to customer behavior in buying products by finding the association or pattern between product sales. The purpose of this study is to obtain association models between items that can be used as a base to arrange recommendations for selling beverage products at Cafe X using a data mining method, namely the association rules with the Apriori algorithm. The transaction data specified for data processing are payment transactions of a total of 3,100 data records. From the analysis results, it is known that there are 8 product association rules that meet the minimum support of 0.01 and the minimum confidence of 0.5, which means that these rules can be used as an alternative basis for determining business strategy at Cafe X. Based on the association information between products, the customer relationship management (CRM) strategy that can be applied at Cafe X is the promotion strategy of product bundling. From a number of recommended bundled products, it is found that there are opportunities to increase sales for four products, they are MD001, TH002, LB001, and TU001.