Is More Social Presence Always Better? Exploring Boundaries to Consumer Acceptance of Robots

Abstract

Frontline service interactions have traditionally been fundamentally social interactions. The automation, or replacement of human employees by technologies, has a profound effect on the nature of service interactions and the customer experience. Retailers are in search of adding a human touch to a technology-mediated service interface in the store, and a well-received tool to this end is humanoid service robots. These human-like robots in the frontline have been found to instill trust and generate better consumer adoption rates (e.g., Van Doorn et al., 2017; Wirtz et al., 2018). The question on how far the initial promising results of employing humanoid service robots in retailing reach, has remained unanswered until present. Apart from the ‘uncanny valley’ effect, whereby eeriness dominates consumer perceptions of a too human-like robot, further boundaries to the generalizability of the positive effects of social presence are yet to be established. Answering Grewal et al.’s (2020) recent call, the present empirical study examines personal and situational factors to shed light on this underexplored dark side of social presence. First, in terms of personal characteristics, consumer’s contamination fear in the COVID-19 pandemic is explored in studying preferences toward human-human or rather human-tech interactions. Second, we explore embarrassing purchase situations in which more social presence is not necessarily better, or even undesirable. Negative feelings of embarrassment occur when people’s desired social identity is disrupted in the presence of other people (Goldfarb et al., 2015; Lee et al., 2020). Think for example of purchasing private products (e.g., condoms; Otterbring et al., 2018). The desire in such cases to avoid negative social judgment can clarify why interpersonal interactions could negatively impact customer outcomes like satisfaction (Dahl et al., 2001). Given technologies’ lack of judgment, we examine whether shoppers in such situations prefer technologies over humans (Pickard et al., 2006).