Leveraging "human-likeness" of robotic service at restaurants

Despite the rise of human-robot interaction research, the mixed findings of human-likeness in consumer evaluation exist. Focusing on the restaurant sector, this research investigates how service robots' varying levels of human-likeness of attributes (i.e., visual, vocal and verbal) influence consumption outcomes (e.g., service encounter evaluation, revisit intentions and positive word of mouth intentions) and the underlying mechanisms through cognition (i.e., perceived credibility) and positive emotion per Appraisal Theory. Drawing on a consumer experiment involving a total of 587 participants, results suggest that humanlike voice emerges as a dominant attribute affecting all three consumption outcomes. Humanlike language style positively affects service encounter evaluation but barely affects the other two outcomes. The significant effect of humanlike voice on three consumption outcomes is only explained by positive emotion whereas the effect of humanlike language style on service encounter evaluation is explained by both cognition (i.e., perceived credibility) and emotion.