FairPlay: Detecting and Deterring Online Customer Misbehavior

Customer misbehavior is a serious and pervasive problem in firm-sponsored social media, yet prior studies provide limited insight into how firms should detect and manage it. To address this gap, we first develop a data science approach to detect customer misbehavior on social media and then devise intervention strategies to deter it. Specifically, we build on natural language processing and deep learning techniques to automatically detect customer misbehavior by mining customers’ social media activities in collaboration with a leading apparel firm. The results show that our algorithmic solution achieves superior performance, improving detection by 7%–9% compared with traditional methods. We then implement two types of intervention policies based on the focus theory of normative conduct that advocates the use of injunctive norms (i.e., a punishment policy) and descriptive norms (i.e., a common identity policy) to restrain customer misbehavior. We conduct field experiments with the firm to validate these policies. The experimental results indicate that punishment considerably reduces customer misbehavior in the short term, but this effect decays over time, whereas common identity has a smaller but more persistent effect on misbehavior reduction. In addition, punishing dysfunctional customers decreases their purchase frequency, whereas imposing a common identity increases it. Interestingly, our results show that combining the two policies effectively alleviates the detrimental effect of punishment, especially in the long run. We examine the heterogeneous treatment effect on novice and experienced customers. Finally, a follow-up field experiment reveals that the disclosure of the use of an artificial intelligence detector improves the effectiveness of the intervention strategies, and this effect is more pronounced for the punishment and combination strategies.