EEEU 2024 conference context; Springer first online: June 18, 2025 (DOI: 10.1007/978-981-96-4116-1_9)

Virtual Influencer Cues and Gen Z Purchase Intention

This study tests Instagram-style virtual influencer content using a 2 x 2 cue experiment with Gen Z respondents in Vietnam. It evaluates central and peripheral cue effects on purchase intention and the moderating role of speciesism.

Conference context: 2024 Springer first online: June 18, 2025 (DOI: 10.1007/978-981-96-4116-1_9) Co-Author

Background

Background and Research Question

Virtual influencers are growing in Vietnam, but evidence on how Gen Z responds to different persuasive cues is still limited. This study tests how cue strength in virtual influencer content changes purchase intention, and whether speciesism changes that response.

Sample

N = 244 valid responses

Design

2 x 2 factorial experiment

Platform Task

Simulated Instagram feed post

Product Context

Headphone endorsement scenario

Virtual influencer

A virtual influencer is a computer-generated social media persona designed to act like a human influencer. It can post content, interact with followers, and endorse products.

Speciesism

Speciesism is a bias that gives higher moral value to humans than non-human entities. In this study, higher speciesism is linked to weaker purchase intention toward virtual influencers.

ELM cue logic

The central cue is argument quality. Peripheral cues are social attractiveness, physical attractiveness, and attitude homophily. The study tests both cue effects and their interaction.

Photoreal virtual influencer visual context (fashion style) — Example context: photoreal virtual influencer profile.

Example context: anime-style virtual influencer profile.

Example context: 3D virtual influencer profile.

Study source: Dai Chung Hy et al., Exploring the effect of social media content of virtual influencers on Generation Z purchase intention, EEEU 2024.

EEEU 2024 one-page infographic preview for virtual influencer study

EEEU 2024 Infographic

One-page visual summary of the virtual influencer trust study, added in a compact format for quick scan.

Open Infographic Download

Method

Method and Instagram Simulation

Participants were randomly assigned to one of four stimulus blocks in a 2 x 2 design: central cue high or low, and peripheral cue high or low. The survey asked respondents to imagine scrolling Instagram and evaluate a virtual influencer post that promoted headphones.

Design2 x 2 factorial

Blocks4 randomized blocks

Block size60, 62, 62, 60

Platform simulationInstagram post + caption

AnalysisRegression + ANOVA

Age profileMean 20.04 (SD 1.78)

Instagram simulation screenshot for high central and high peripheral cue condition — Questionnaire simulation structure across four cue conditions; Block 1 shows the high-central and high-peripheral condition.

Instagram simulation screenshot for high central and low peripheral cue condition — Questionnaire simulation structure across four cue conditions; Block 1 shows the high-central and high-peripheral condition.

Manipulation checks confirmed cue separation: peripheral cue t(242) = -5.44, p < .001 and central cue t(242) = -5.05, p < .001.

Results

Regression results show that social attractiveness, physical attractiveness, and attitude homophily were positive predictors of purchase intention. Argument quality was significant but negative in this model (adjusted R² = 0.30).

Reported regression coefficients and ANOVA test summary from the virtual influencer experiment with Generation Z participants — Reported regression coefficients and ANOVA outputs for the study sample (N = 244).

Peripheral cue mean difference and speciesism moderation pattern from the virtual influencer experiment — Peripheral-cue effect and speciesism moderation pattern from the reported analysis.

Main effect conclusion

Peripheral cues showed an independent positive effect on purchase intention, F(1, 240) = 7.284, p = .007, while central cue main effect was not significant.

Interaction conclusion

The central by peripheral interaction was not significant, F(1, 240) = 0.08, p = .779, so the combined interaction hypothesis was not supported.

Speciesism conclusion

High-speciesism participants remained low in purchase intention even when peripheral cues were stronger, while low-speciesism participants responded more positively.

Managerial Implications

Cue sequencing rule for campaigns

Design strong peripheral cues to capture attention, but pair them with clear and credible product claims to sustain trust in later campaign stages.

Segment strategy should account for speciesism. High-speciesism segments may require additional human reassurance signals rather than aesthetics alone.

Build peripheral cues deliberately

Increase social and physical appeal with culturally relevant styling and homophily cues for Gen Z audiences.

Protect message credibility

Maintain argument quality in captions and product claims, especially for higher-involvement categories.

Segment by speciesism profile

For audiences with higher speciesism, combine virtual endorsers with stronger humanized cues and reassurance.

Execution guardrail

Use clear disclosure that the endorser is virtual, and avoid manipulative framing in youth-focused campaigns.

Back to All Research