Overview
This series asks when virtual influencers—computer-generated social endorsers—create useful influence and when artificiality becomes a liability. Comparative evidence across 210 independent experimental studies from 110 papers, 643 effect sizes, and 91,058 consumers finds higher novelty but no reliable overall engagement advantage and average disadvantages in credibility, attitudes, attractiveness, and behavioral intention. The chapters explain how anthropomorphism shifts evaluation between novelty and credibility, how autonomy, brand positioning, product newness, sensory demands, and message appeal shape fit, and why novelty may weaken while credibility matters more in later study periods. Control and scalability solve production problems, not necessarily legitimacy. Use the findings as conditional design and testing guidance, not a universal forecast: period comparisons are not a same-person lifecycle, and long-term trust, sales, ROI, and broader cultural effects remain unsettled.
Learning outcomes
- Virtual influencers can be more novel without producing a reliable overall engagement advantage.
- In pooled comparisons, virtual influencers average lower on credibility, brand and ad attitudes, attractiveness, and behavioral intention.
- Anthropomorphism changes the route: stylized designs strengthen novelty, while human-like designs narrow—but do not erase—the credibility gap.
- Fit depends on autonomy, brand positioning, product newness, sensory demands, and message appeal; unsupported interactions should remain flexible.
- Control and scalability support production, not proof of trust, profit, long-term value, or legitimacy.
Keywords
Preview every chapter · Unlock full text, audio, and video from chapter 4
PlansFAQ
Do virtual influencers improve marketing performance?
Not consistently: virtual influencers are more novel on average, show no reliable overall engagement advantage, and are lower on several persuasion-oriented outcomes in pooled comparisons. The pattern spans varied products, audiences, platforms, and comparison conditions, so it is not a forecast for every campaign.
When do virtual influencers fit a campaign?
They fit best when distinctiveness is valuable, or when their design and context make a credibility-oriented role plausible. Stylized personas can support novelty; human-like personas need coherent autonomy, brand, product, message, and governance. This is a testable hypothesis, not a guaranteed lift.
Does a more human-like virtual influencer become more persuasive?
Not universally: high anthropomorphism narrows the comparative credibility disadvantage, while low anthropomorphism produces stronger novelty, and neither guarantees persuasion. Human likeness changes the audience’s likely evaluation route; it does not create lived experience.
Will virtual-influencer novelty fade over time?
It may weaken as the format becomes familiar, but the evidence compares study periods rather than following the same consumers or a specific account. Recent studies show weaker novelty and a less negative credibility path, so maturation is a planning hypothesis, not a fixed lifecycle.
Can control and scalability substitute for trust?
No: control and scalability solve production and coordination problems, not trust, legitimacy, or accountability. A controlled identity still needs clear disclosure, bounded autonomy, responsible claim review, and outcome testing; efficiency is not evidence of profit or long-term value.