Incorporating social learning into the optimal return and pricing decisions of online retailers

The biggest drawback of online shopping lies in consumers’ uncertainty about the product. Before making a purchase decision, consumers learn about the product from many sources which helps them infer whether a product meets their preferences, i.e., social learning. Social learning constitutes a positive or a negative signal about the product resulting in positive-type consumers and negative-type consumers. Also, many retailers offer money-back guarantees (MBGs) to reduce consumers’ mismatch risk. We analyze a retailer’s pricing and return policy decisions in the presence of social learning. Under both no returns and MBG, we find that when signal accuracy is low, it is optimal for the retailer to sell to both positive and negative-type consumers. When signal accuracy is high, it is optimal for the retailer to sell to only positive-type consumers. In addition, if no returns are allowed, social learning hurts the retailer when signal accuracy is low while social learning benefits the retailer when signal accuracy is high. If a MBG is offered, social learning always benefits the retailer. When consumers are heterogeneous in their belief about signal accuracy, if no returns are allowed, social learning benefits the retailer for a small match probability, whereas social learning hurts the retailer for a large match probability. Similar to the case of homogenous consumers, social learning benefits the retailer in almost the whole range of problem parameters if a MBG is offered. Moreover, we find that social learning can be a driving factor to the ubiquity of MBGs regardless of the heterogeneity of consumers. We also analyze the effects of quality effort investment, consumers’ full learning, consumers’ bounded rationality, and high salvage value on the retailer’s pricing strategy.