Service encounters with virtual agents: an examination of perceived humanness as a source of customer satisfaction

Purpose: Firms have begun to introduce virtual agents (VAs) in service encounters, both in online and offline environments. Such VAs typically resemble human frontline employees in several ways (e.g. the VAs may have a gender and a name), which indicates the presence of an assumption by VA designers – and by firms that employ them – that VA humanness is a positively charged characteristic. This study aims to address this assumption by examining antecedents to perceived humanness in terms of attribution of agency, emotionality and morality, and the impact of perceived humanness on customer satisfaction. Design/methodology/approach: A questionnaire was distributed online to participants who had been interacting with existing VAs, and they were asked to focus on one of them for this study. The questionnaire comprised measures of antecedents to perceived humanness of VAs, perceived humanness per se and customer satisfaction. A structural equation modeling approach was used to assess associations between the variables. Findings: Attributions of agency, emotionality and morality to VAs contributed positively to the perceived humanness of the VAs, and perceived humanness was positively associated with customer satisfaction. Research limitations/implications: Additional humanness capabilities should be explored in further research. Practical implications: Firms using VAs in service encounters should make attempts to maximize perceived VA humanness, and this study shows that it may be beneficial if such attempts comprise signals that VAs have agency, emotionality and morality. Originality/value: By examining VAs in terms of a set of fundamental human capabilities, the present study contributes to existing research on human–VA service encounters, which to date has focused on more superficial VA characteristics (such as if the VA has a face and gender).