The impact of providing non-human identity cues about sales agents on consumer responses: the role of social presence and speciesism activation

De Cicco, Roberta, Elmashhara, Maher Georges ORCID: https://orcid.org/0000-0002-1576-560X, Silva, Susana and Hammerschmidt, Maik (2025) The impact of providing non-human identity cues about sales agents on consumer responses: the role of social presence and speciesism activation. European Journal of Marketing, 59 (13). pp. 55-84. ISSN 0309-0566

Preview

Published Version Available under License Creative Commons Attribution. Download (723kB) | Preview

Abstract

Purpose: This work investigates how different strategies for providing cues about the non-human identity of a sales agent influence consumers’ perceptions and purchase-related outcomes, and how a social interaction style shapes these responses. Additionally, the authors explore the role of consumers’ speciesism against non-human entities in eliciting unfavourable responses to the disclosure of the agent’s artificial nature. Design/methodology/approach: Three experimental studies were conducted using real chatbot interactions. Study 1 investigates how non-human identity cues impact consumer trust and, subsequently, attitude towards the firm and intention to purchase the product offered. Study 2 tests these effects across different levels of social presence. Study 3 examines consumer responses to different non-human identity disclosure strategies, considering speciesism’s moderating role. Findings: Study 1 proves that disclosing (vs not disclosing) the artificial nature of a sales agent leads to a decline in trust towards the firm, which in turn negatively influences both attitude towards the firm and purchase intention. This finding reveals discrimination against disclosed (vs non-disclosed) artificial sales agents despite identical, flawless performance. However, Study 2 proves that the negative effects vanish when perceived social presence is high. Study 3 underlines that high speciesism leads to a trust decline if non-human identity cues are presented during the interaction but not if presented earlier in the journey before the interaction. Research limitations/implications: The study highlights the negative effects of disclosure on important, firm-related outcomes. These insights advance current literature by showing that disclosing cues about the non-human nature of a sales agent can undermine psychological and behavioural responses–even when the disclosed agent performs just as effectively as its undisclosed counterpart. This result is noteworthy, as most prior research has linked aversive reactions to artificial agents with situations in which algorithms underperform, whereas this study examines agents that function flawlessly. Furthermore, the study reveals that these adverse effects are driven by speciesism–prejudices against non-human entities–offering a novel explanation for consumers’ negative responses. Practical implications: The findings stress that transparency about the artificial nature of sales agents is penalised by customers and comes at a high cost for business-relevant outcomes. However, by transforming an artificial agent into a social actor through subtle design modifications, firms can overcome the unfavourable prejudice against artificial agents. By creating a social appearance, firms can harness the potential of automated sales services–even when disclosure of the agent’s artificial identity is required. As firms may soon be obliged to disclose the artificial identity of their sales agents, the critical question shifts from whether to disclose to how to disclose in order to mitigate negative consequences. Finally, we offer guidance on targeting the right consumers with artificial agents–specifically, those with lower levels of speciesism-related prejudices. Originality/value: This work addresses pressing issues for managers concerned with the implementation of artificial sales agents. Results extend knowledge on speciesism towards digital agents, inform which consumers are particularly prone to respond negatively to such agents, and present levers for designing chat-based social interactions that prevent non-human-related prejudices that could undermine the effectiveness of conversational technologies.