AI automation and Retailer Regret in Supply Chains

Abstract

In this paper, we explore how Artificial Intelligence (AI) delivers values in supply chains. In particular, we study the effects and implications associated with AI automation of order decision in a decentralized supply chain comprising one supplier, one emotional (particularly, regretful) retailer, and a wholesale price contract that governs the transactions between them. In this context, regret is defined as a cognitive bias that essentially describes retailers who behave as though considering inventory error, the deviation between the order quantity and realized demand, when making an ex ante inventory decision. We find that, if profit margins of the supply chain are high, regret bias drives the retailer to decline the supplier’s contract, whereas,if profit margins are low, regret drives retailers to order more from the supplier. As a result, although the automation of retailer decision unequivocally leads to a higher expected profit for a retailer that operates in a centralized vacuum, it nevertheless can be a negative force for a decentralized supply chain with either high or low profit margins. Perhaps more interestingly, as a retailer’s decision becomes automatic, it is not necessarily destined to earn a higher expected profit. In the extreme, a lose-lose outcome can prevail in which automation can potentially leave both the newsvendor and supplier worse off.