Analysis of consumer characteristics on retail business with clustering analysis method and association rule for selling improvement strategy recommendations

Khasanah, Annisa Uswatun
Baihaqie, Muhammad Rafly Qowi

OPSI
Vol 17 No 1 (2024): ISSN 1693-2102

In the highly competitive retail industry, companies must continually innovate and develop unique business strategies to enhance their sales performance. The ABC Store, a mini market in Yogyakarta, has experienced fluctuating sales over the past year, failing to meet its targets. This study aims to analyze consumer purchasing behavior at the ABC Store and provide strategic recommendations to boost sales. The data analyzed in this study comprises three months of transaction records. The methods used include Association Rule - Market Basket Analysis (AR-MBA) with the FP-Growth algorithm and Clustering Analysis with K-Means. The clustering analysis identified four distinct customer segments: Mid-Morning Moderates, Diverse Afternoon Buyers, Evening Moderates, and High-Value Customers. Cluster 2, comprising Diverse Afternoon Buyers, was selected for AR analysis due to its relatively high transaction value and the variety of products purchased, indicating its potential to evolve into a High-Value Customers cluster. The analysis yielded 104 rules. The findings can inform marketing strategies to increase sales, including product bundling and customer loyalty programs such as a point system.In the highly competitive retail industry, companies must continually innovate and develop unique business strategies to enhance their sales performance. The ABC Store, a mini market in Yogyakarta, has experienced fluctuating sales over the past year, failing to meet its targets. This study aims to analyze consumer purchasing behavior at the ABC Store and provide strategic recommendations to boost sales. The data analyzed in this study comprises three months of transaction records. The methods used include Association Rule - Market Basket Analysis (AR-MBA) with the FP-Growth algorithm and Clustering Analysis with K-Means. The clustering analysis identified four distinct customer segments: Mid-Morning Moderates, Diverse Afternoon Buyers, Evening Moderates, and High-Value Customers. Cluster 2, comprising Diverse Afternoon Buyers, was selected for AR analysis due to its relatively high transaction value and the variety of products purchased, indicating its potential to evolve into a High-Value Customers cluster. The analysis yielded 104 rules. The findings can inform marketing strategies to increase sales, including product bundling and customer loyalty programs such as a point system.