Who did what and when? The effect of chatbots’ service recovery on customer satisfaction and revisit intention

Purpose This study aims to explore the recovery performances of chatbots (vs human employees) and help firms use chatbots to carry out effective service recovery. Design/methodology/approach Two experiments were conducted to test the proposed hypotheses. Findings The results show that compared with human employees’ recovery, chatbots’ recovery leads to lower customer satisfaction and revisit intention. This effect is more significant for symbolic recovery instead of economic recovery. Perceived distributive and interactional justice mediate the interaction effect of recovery provider and recovery strategy on recovery performance. Using immediate recovery rather than delayed recovery can attenuate chatbots’ poor performances in symbolic recovery. Originality/value This study enriches the chatbot research and the service recovery literature by deploying chatbots into the service recovery setting. Using an integrated theoretical model including recovery strategy and recovery timing, this study provides substantive insight into how firms can enhance chatbots’ recovery performances.