Customer returns pose a big problem for retailers selling online due to high costs and CO2 emissions. This paper introduces a new concept to handle online returns, the customer-to-customer (C2C) return logistics. The idea behind the C2C concept is to deliver returning items straight to the next customer, skipping retailers’ warehouse. To incentivize customers to purchase C2C returning items, retailers can promote returning items on their webshop with a discount. We build the mathematical models behind the C2C concept to determine how much discount to offer, ensuring that enough customers are triggered to purchase C2C returning items and the expected total profit of the retailer is maximized. Our first model, the base model (BM), is a customer-based formulation of the problem and provides an easy-to-implement constant-discount-level policy. Our second model formulates the real-life problem as a Markov decision process (MDP). Since our MDP suffers from the curse of dimensionality, we resort to simulation optimization (SO) and reinforcement learning (RL) methods to obtain reasonably good solutions. We apply our methods using data collected from a Dutch fashion retailer. Furthermore, we provide extensive numerical experiments to claim generality. Our results indicate that the constant-discount-level policy obtained with the BM performs well in terms of expected profit compared to SO and RL. With the C2C concept, significant benefits can be achieved both in terms of expected profit and return rate. Even in cases where the cost-effectiveness of the C2C return program is not pronounced, the proportion of customer-to-warehouse returns to total demand gets lower. Hence, the system can be defined as more environmentally friendly. The C2C concept can help retailers in addressing the online return problem financially and adhering to the growing need for corporate social responsibility from the last decade.
|Number of pages||34|
|Publication status||Published - 2022|