Stock Management Strategy Based on Product Popularity Cluster Analysis

Abstract

Improving customer service in terms of stock availability remains a challenge. Supermarkets often face issues such as stock shortages and overstocking, leading to customer dissatisfaction. To address these challenges, research was conducted on stock management based on cluster analysis of product popularity, aiming to optimize marketing strategies related to product availability. This study employed the K-means clustering algorithm using Google Colaboratory tools. The clustering results revealed that Cluster 1 represents popular products, characterized by a higher average number of products sold and relatively higher customer satisfaction levels compared to Clusters 0 and 2. Cluster 0 represents moderately popular products, while Cluster 2 encompasses less popular products. Based on the visualization, it was observed that product popularity in each city varies: electronic products are most popular in Mandalay and Naypyidaw, while health and beauty products are less popular or fall under Cluster 2 in Naypyidaw. However, health and beauty products are highly popular and record the highest sales in Yangon.