Automated social presence in artificial agents: A conventional and qualitative comparative meta-analysis
(2026) Journal of the Academy of Marketing Science — (2026)
Files
2026_JAMS_Juquelier Hazée et al (postprint).pdf
Postprint version
Embargo Access until 2027-07-06 - Adobe PDF
- 4.61 MB
Details
- Authors
- Abstract
- As firms increasingly deploy artificial agents with social capabilities, the effectiveness of automated social presence (ASP), defined as agents’ ability to evoke the perception of interacting with a socially present other, has become a top research priority. However, empirical findings remain mixed and fragmented. This study meta-analytically integrates 165 studies (N = 54,917) to clarify ASP’s effects, underlying mechanisms, boundary conditions, and antecedents. The results reveal that ASP is strongly and positively associated with customer satisfaction, engagement, and loyalty, and shows a weaker relationship with psychological well-being. A mediation analysis shows that social cognition and psychological ownership drive distinct outcome pathways. Moreover, ASP’s effects weaken in emotionally charged and long-term encounters. Finally, a crisp-set qualitative comparative meta-analysis, using a novel two-dimensional taxonomy of artificial agent design characteristics, reveals how multiple configurations of capability (knowledge domain, interaction modality, interaction flexibility) and identity features (appearance, name, female gender) interact in shaping ASP. These findings refine theoretical understanding of ASP and offer insights for designing artificial agents that evoke ASP
- Affiliations
Citations
Juquelier, A., Hazée, S., Van Vaerenbergh, Y., & Poncin, I. (2026). Automated social presence in artificial agents: A conventional and qualitative comparative meta-analysis. Journal of the Academy of Marketing Science. Accepted/in-press. https://doi.org/10.1007/s11747-026-01177-x (Original work published 2026)