How should robots apologize to restore satisfaction after a service failure?
(A2023-114513)

AUTHORS

Salma André, EMLYON Business School, University Jean Moulin Lyon 3, iaelyon School of Management, UR Magellan; Margherita Pagani, Skema Business school

ABSTRACT

Companies are increasingly using chatbots to connect with their customers at different stages of the customer journey. Despite the opportunities brought by this technology, its deployment could result in service failures. Considering the salience of emotional recovery to restore customer satisfaction following a service failure, this research examines the effect of the chatbot apology on post-failure satisfaction. We build on mind perception theory and justice theory to propose that the effectiveness of the chatbot apology is contingent on its perceived feeling abilities (affective and physiological abilities) and perceived agency (cognitive abilities and self-control). Results from an experimental study involving a real interaction with a chatbot show that the chatbot apology leads to improved customer satisfaction after a service failure only when it was perceived with high feeling abilities. This pattern of effect was found only when the robot was perceived with low agency relative to high agency. This research has important theoretical and practical implications about customer reaction to service failures caused by robots.