Artificial intelligence has revolutionized retail by enhancing business decision-making. This research applies the RFM (Recency, Frequency, Monetary) framework for customer segmentation, promoting sustainable consumer behaviour and eco-friendly products. Mobility issues, such as efficient goods movement and customer access, are also pivotal in sustainable retail supply chains. A systematic literature review (SLR) and Python-based clustering techniques (K-Means, hierarchical, DBSCAN) are employed to analyse a four-year dataset of customer data. The SLR identified six key areas from 71 articles. Clustering results varied: RFM binning found four clusters, K-Means and Mean Shift found three, and hierarchical and DBSCAN found two. The study emphasizes a data-centric retail strategy and the transformative impact of machine learning on customer engagement.

Background: Globally, in the Fast-Moving Consumer Goods (FMCG) industry, excess inventory results from the bullwhip effect. Earlier, barcode-based two-bin systems were limited by manual scanning; hence, a more responsive system is needed to align the inventory with real-time demand. Prior studies have predominantly concentrated on mitigating demand fluctuations and employed comparatively low-efficiency systems, hindering excess inventory (EI) reduction. Methods: This study proposes identifying research gaps, considering the distributor-manufacturer relationship, and developing an RFID-based modified two-bin system and mathematical model to reduce EI and control over manufacturers’ excessive cost. Results: This study tested through Python-based simulation using historical data from an FMCG manufacturer, and the proposed model achieved a reduction in 67% EI and 73% month-wise holding costs. Moreover, the integration of the Artificial Bee Colony algorithm optimizes rework rates within budget, including reworking shop-floor and holding costs, contributing to a monthly excessive cost reduction of 34–48%, alongside a corresponding 41–44% cumulative excessive cost reduction. Conclusions: Bringing significant implications on digitalized SCM, this study offers a practical and scalable solution for perishable FMCG items facing demand variability and budget constraints. Collectively, this novel perspective bridges research gaps and motivates future research for embedding trend-aligned parameters, enhancing the model’s performance through diverse SCM contexts like safety stock and backorder cost optimization.