Inventory Management of Perishable Foods in Omnichannel Retail Using Predictive Analytics: A Case Study

Managing perishable seafood sales in an omnichannel retail environment requires balancing customer expectations for freshness, quality, and affordability with the retailer’s goals of minimizing waste, ensuring inventory turnover, and maintaining supply chain efficiency for both locally sourced and imported products. Challenges in omnichannel is heightened by the need to synchronize inventory and demand across physical and online platforms, address faster delivery requirements, and navigate constraints like real-time stock visibility, temperature-controlled logistics, and diverse customer preferences, especially for perishables. Despite the growing relevance of omnichannel strategies, there is limited literature on optimizing perishable inventory in this context. This study bridges the gap by analyzing omnichannel sales and inventory distribution of a retail chain using predictive analytics. Gradient Boosting (GB) emerged as the best-performing model for demand forecasting, achieving the lowest MAE (1.684) and MSE (7.76) on average. A mathematical model based on predicted demand is suggested for efficient inventory control. Furthermore, our analysis suggests that segmenting forecasts for online and in-store sales, accounting for day-of-week trends, can optimize inventory management, reduce waste, ensure product availability, and adhere to food safety standards, particularly for perishable seafood items.