A few years ago, brands used virtual influencers as solo acts: one CGI persona, one campaign, one product. That pattern is thinning out. Samsung put Lil Miquela next to Millie Bobby Brown for #TeamGalaxy. Calvin Klein paired Miquela with Bella Hadid. Magnum matched Imma with singer Weilong Song in China. The new default is the pairing, not the persona alone, and the pairing changes what the campaign is actually selling. This piece looks at what a virtual influencer collaboration is now, why co-endorser choice tends to decide the outcome, and where these campaigns still fall flat.
The Shift From Standalone Endorsements to Collaborations
Early virtual influencer deals were straightforward endorsements. A persona wore the jacket, held the phone, tagged the brand. Buyers liked the control, and audiences liked the novelty, but the story rarely traveled past a single post.
A recent multiple-case study of five virtual influencers (Miquela, Imma, Noonoouri, Lu do Magalu, and Knox Frost) traces the drift toward collaboration formats. Brands still hire virtual influencers for novelty and risk control, but they increasingly stage them with a co-endorser: another persona, a human celebrity, or a supporting cast of creators. The reason is not mystery. A solo synthetic post reads as an ad; a duet reads as a relationship, and relationships hold attention longer.
That drift is now the shape of most large virtual influencer marketing programs, not the exception.
The Two Formats That Dominate
Virtual influencer collaborations fall into two working formats, each with a different job.
The first is VI paired with human. Miquela with Bella Hadid, Imma with Weilong Song, or a supermodel photographed alongside a CGI persona in the same frame. This format buys borrowed humanity: the human anchors the visual so the audience does not have to decide whether the persona is “real.” The campaign is about proximity.
The second is VI paired with VI, or VI paired with a fictional universe. Cross-persona features, joint music drops, K/DA style ensembles, or crossovers with anime IP. This format buys shared canon: fans who follow one character get pulled into the world of the other, and the story compounds. The campaign is about lore.
Both formats sit inside a wider trend of hybrid campaigns, where a human spokesperson and a virtual avatar assistant work as a pair. Reported engagement lifts are real, but the shape of the lift depends heavily on which co-endorser you pick.
Why Co-Endorser Choice Is the Whole Ballgame
A 2025 study in Journal of Retailing and Consumer Services ran three experiments on how audiences respond to virtual influencer collaborations. The finding, in plain terms: success depends less on whether a co-endorser is present and more on how the pairing is constructed.
Two variables carried most of the effect. The first was perceived distance between the virtual influencer and the co-endorser (whether they read as close friends or distant acquaintances in the shot). The second was gender congruence (same or opposite gender pairings). Neither variable is neutral. Closer relational framing generally produced stronger consumer responses, and gender congruence interacted with product category rather than acting on its own.
The practical takeaway is quiet but useful. When Calvin Klein put Miquela and Bella Hadid in the same frame, the visual grammar (touch, proximity, eye contact) was doing as much work as the casting. When brands drop a persona next to a celebrity with no relational cue, the pairing sits inert on the page.
The Emotional Layer: Why Fans Show Up for Collabs
The audience side of the equation is stranger than the brand side. People who follow a virtual influencer often feel a parasocial bond, and collaborations can intensify that bond by giving fans a “crossover event” to talk about. This is a real effect, but the direction of causation is easy to overstate.
A 2025 Current Psychology paper on parasocial relationships with influencers and loneliness found a measurable correlation: people who report higher loneliness tend to form stronger parasocial bonds with online personalities. What the paper does not show is that following virtual influencers causes loneliness, or that lonely people are the only ones forming these bonds. The link is correlational, and the mediating role of real friendships matters a lot.
For collaborations, the honest reading is that crossover moments amplify existing bonds rather than create them. Fans who already care about a persona show up for the collab; the campaign gives them a reason to gather. Framing a virtual influencer collaboration as an “antidote to loneliness” overreaches, and it will not survive the next round of research. Framing it as a fandom event that strengthens attachment already in place is closer to what the data supports.
Where Collaborations Fall Flat
Not every pairing lands. Three failure modes recur.
The first is aesthetic mismatch. A photorealistic 3D persona placed next to a hand-drawn anime character reads as two different worlds sharing an ad, not two figures sharing a moment. The audience registers the seam.
The second is narrative arm’s length. A collaboration with no shared setup, no story beat, no callback to either persona’s canon feels transactional. Fans of both characters can tell when neither team was allowed to write anything.
The third is undisclosed synthetic content in a mixed cast. When a human celebrity, a virtual influencer, and a smaller creator are all in the same campaign, viewers sometimes assume the smaller creators are also CGI, or that the human celebrity’s likeness has been AI-modified. The 2026 EU AI Act Article 50 disclosure requirements make this a working problem, not a hypothetical.
Brands treating collaborations as “just a bigger post” tend to hit at least one of these. Brands treating them as coproductions, with shared writing rooms and a real reason for both parties to appear together, tend to avoid them.
Practical Guardrails for Brands
A few things separate the collaborations that ship well from the ones that get ratio’d. First, write the relationship before you shoot the frame. Decide whether the personas are meeting, working, playing, or performing together, and let that decide the composition. Second, respect the co-endorser’s own canon. Pulling Miquela into a Samsung set works because her existing lore already includes tech and fashion; putting a character with a stated indie-music identity into a fast-fashion drop confuses their fans more than it reaches new ones. Third, disclose. The upside of a virtual influencer collaboration lives in trust, and trust survives a #PoweredByAI tag more easily than it survives a fan discovering the truth six weeks later.
For a broader view of how these pairings fit inside campaign budgets and channel plans, our overview of virtual influencer marketing walks through the surrounding decisions. If you are still choosing which persona to pair with your brand or celebrity, the top virtual influencers rundown can shorten the shortlist.
Virtual influencer collaboration is not the answer to whether synthetic characters “work.” It is the answer to how they scale past the first campaign. The persona alone is a novelty; the pairing is a program, and the pairing is where the interesting work lives.
Frequently Asked Questions
What counts as a virtual influencer collaboration?
Any campaign where a virtual influencer appears alongside a second endorser: a human celebrity, another virtual influencer, or a fictional character from a linked IP. Solo posts sponsored by a brand are endorsements, not collaborations.
Do virtual influencer collaborations get better engagement than solo posts?
Often, yes. Reported case studies (for example, Miquela with Calvin Klein) show meaningful engagement lifts. But 2025 experimental work suggests the lift depends on how the pairing is staged, not on the presence of a co-endorser alone.
Should the virtual influencer be paired with a same-gender or opposite-gender co-endorser?
The 2025 study on co-endorser type found gender congruence interacts with product category rather than acting as a universal rule. Same-gender pairings tend to work for personal-care and fashion; opposite-gender pairings can work for lifestyle and entertainment, if the relational framing supports it.
Do collaborations require disclosure that one participant is synthetic?
In most jurisdictions in 2026, yes. The EU AI Act Article 50 and updated FTC guidance both push toward clear labeling. #PoweredByAI, “created with AI,” or an explicit persona bio note is the safe default, and audience research suggests disclosure does not meaningfully hurt engagement.