Uncovering Business Accounts Among Retail Shoppers: Insights from Supermarket Transactions Using an Intelligent Decision Support System

Customer segmentation is crucial to retail operations, pivotal in marketing strategies, and improving operational efficiency. This study proposes an innovative approach to optimizing customer segmentation and enhancing business efficiency. The primary objective is to classify customers currently shopping in supermarkets but exhibit behaviors indicative of potential interest in preferring the wholesale channel. Although the propensity of retail customers to buy specific products and their probability of churn have been well documented by previous marketing analytics research, less is known about service-oriented customer segmentation. By leveraging customer behavior and transactional sales-related data from a retail grocer serving both retail and wholesale channels, this research aims to identify retail customers exhibiting wholesale behavior patterns. Various unsupervised classification algorithms were employed during the estimation process. Among them, the KMeans algorithm, known for its usage in segmentation models, was chosen as the primary classification model. Implementing this strategy has the potential to streamline customer targeting efforts, ultimately maximizing the effectiveness of directing individuals toward the most relevant retail outlet. By discovering these patterns, we aim to direct customers toward the retailer’s most efficient service channel, streamlining customer targeting efforts and enhancing operational efficiency.