Is figurative or literal language more effective in the service encounter context? The current research reveals that literal language used by a service provider might be more congruent with conversational norms, thus leading to better customer evaluation. Most importantly, the effectiveness of language styles (literal vs. figurative) is contingent on a service agent type (human vs. robot vs. kiosk). Specifically, consumers respond more favorably to human service agents who use literal (vs. figurative) language, and due to the notion of anthropomorphism such an effect extends to service robots. However, the language style effect is not observed among service kiosks as they lack humanlike features. Furthermore, perceived credibility is the underlying mechanism explaining the language style effect among service agents.