How should voice assistants be heard? The mitigating effect of verbal and vocal warmth in voice assistant service failure
Bo Huang, Sylvain Sénécal
Abstract
Voice assistants have become increasingly popular touchpoints in AI-infused service encounters in the hospitality industry. Although we have seen a growing body of research in this area, little attention has been paid to specific service failures involving exclusively voice interactions. Drawing from the Computers As Social Actors (CASA) research paradigm and the Stereotype Content Model, this research explores how warmth can mitigate the negative consequences of service failure involving voice assistants. In two experiments using both physiological (EDA) and psychological measures, we show that the perception of warmth improves consumers’ emotional reactions and increases re-patronage intention following a negative service outcome. We also found that the optimal voice to be used for a voice assistant is a dynamic speech style combined with emotionally expressive and warm verbal content. These findings contribute to the knowledge of voice-based smart service interaction and provide insight into how to mitigate the negative consequences of service failure involving voice assistants.