Measuring consumer-perceived humanness of online organizational agents

Abstract

As organizations increasingly adopt artificial intelligence (AI) systems for their online communication with publics, conceptualizing and operationalizing perceived humanness presents an important challenge to understanding how consumers and other publics respond to both AI and human agents in live chats with organizations. This study develops a measure of consumer-perceived humanness derived from three related constructs: conversational human voice, anthropomorphism, and social presence. The new measure was tested for reliability and validity, including tests of criterion validity predicting how it would relate to perceived relational investment and trust as relational outcomes. Two online studies (N₁ = 172, N₂ = 375) were conducted in which participants interacted with either human or machine online chat agents. Greater perceived humanness of chat agents was found to correlate with greater perceived relational investment and trust in the organization, with perceived relational investment mediating the relationship between perceived humanness and trust. Perceived humanness can be utilized subsequently to assess both human and machine agents. Its corollary, perceived machineness, was also identified as a separate factor to be taken into consideration in future research and theory building.