An Industry-Agnostic Approach For The Prediction Of Return Shipments

Return shipments are a major problem in E-commerce not only economically for the companies selling the products, but also for the environment. We introduce a universally applicable Decision Support System (DSS) for the prediction of product returns. Most of the prediction algorithms used in the research were developed on datasets with high return rates and are product feature centric. However, these algorithms work best for fast-moving products with many features that can be found in the fashion industry. We tackle these challenges by using Design Science Research (DSR) to develop a prediction mechanism based on the current shopping cart of a customer. Using a dataset from a German technical wholesale company, we validate and demonstrate our approach. Thus, we identify consumption patterns within fulfillment datasets containing a low product return rate at sufficient accuracy. This allows supplying a broader industry with prediction algorithms that have low return rates.