A Conceptual Approach for an AI-Based Recommendation System for Handling Returns in Fashion E-Commerce

The benefits of fashion E-commerce for customers are undisputed. Ordering clothes is easy, fast and usually free of delivery and return charges. Fashion E-commerce is moving the changing room from the brick-and-mortar store to the consumer’s living room. These efforts, and the associated ever-increasing E-commerce sales, are accompanied by a consumer return behaviour that has high environmental and economic costs. Most studies in this area focus on the prevention of returns. However, many returns cannot be avoided, e.g. those based on selection orders or quality problems. This is the starting point for this study, which investigates how AI can be used to improve reverse logistics from an environmental and economic perspective, based on return forecasts. To this end, a literature review of existing approaches is conducted, followed by a detailed concept of an AI-based recommendation system for the best possible further processing of returns, with the aim of routing the returned product to a suitable sales channel as quickly as possible.