Service outcome valence is a prominent contextual factor that has been under-researched in studies of consumer response to automated service-givers. We investigate whether consumers respond differently to chatbot and human customer service when faced with positive versus negative service outcomes. We also explore the role of perceived empathy as a potential mediator in the focal relationship. 714 individuals participated in three studies based on the vignette method. Study participants reported significantly lower satisfaction, repatronage intentions, recommendation acceptance, and recommending the provider to others following interactions with a chatbot compared with a human agent in both positive and negative service outcome conditions. The effect was fully mediated by the service-giver's perceived empathy. Increasing a chatbot's perceived empathy via more empathetic communication (but not human-like appearance) improved consumer evaluations of chatbot service relative to a less empathetic chatbot configuration and matched evaluations of the human agent. A fourth study involved a quasi-experiment in which 100 individuals interacted with ChatGPT to obtain medical advice. This study corroborated our conclusions about the impact of perceived empathy on consumers' service experience and provider ratings.