AirOps counted 3.5 billion citations across AI answers. Eleven months, August 2025 through June 2026, tracking which sources six AI engines actually pull from when they answer real buyer questions. Two numbers tell this story:
Creator and social citations: up 140%. No other source category even doubled.
The next closest, reviews, grew 38%.
Brand-owned pages: down 10%.
I'm seeing this in citation logs weekly lately. My own posts show up in AI answers again and again. So does the content my clients' creators make. It's very cool to see research put to a trend I've been watching firsthand. Beca Ramón and the AirOps research team just did exactly that, and the numbers are actually bigger than I would have guessed.
First, let's kill the word "influencer" (please 🙏)
Because when I say invest in creators, I can already hear the objection. Dance trends. Sponsored smoothie posts. Accounts with 800K followers, half of them bought, none of them buyers.
That is not what's getting cited.
What's getting cited is the sales leader who breaks down a real deal cycle. The engineer who films herself actually implementing the tool. The practitioner with fifteen years in the field who finally started writing it down. People with scar tissue, not ring lights.
The AI engines can't be flattered and can't be bought. They pull the piece of content that best answers the question a buyer just asked. Vanity metrics don't survive that filter. Expertise does.
Tyler Denk, CEO of beehiiv (and all-around stellar human), said it perfectly last week:
"The age of the 'influencer' is over. The age of the expert with 10+ years of experience, sharing it online for the first time, is just beginning."
Couldn't agree more. The AirOps data is what that shift looks like measured at scale.
Trust is moving
For years that was the SEO play: own the page that answers the question. Build the resource center. Rank the comparison page. Control the answer with content you control. When AI search arrived, so many folks ported the same playbook straight over.
But now what a brand says about itself carries less weight every month. What other people show, real demos, real walkthroughs, firsthand takes, carries more. Down 10% against up 140% is not noise, this is fundamentally trust moving.
The money already knows. Creator spend in the US grew from $13.9B in 2021 to $44B in 2026, nearly four times the growth rate of media spend overall last year. But most of those dollars still fund creators the old way: a campaign here, a sponsored flight there, measured in impressions. Meanwhile the citations pile up in a channel almost nobody is measuring.
The engines cite content, not channels
This is the finding that should change how you evaluate every creator you work with.
On YouTube, 99% of citations point to one specific video. Not the channel. On LinkedIn, published posts take 89% of citations. Personal profiles and company pages combined take 7%.
Seven percent. A decade of consistent posting, a six-figure following, a beautifully branded channel. The engines barely glance at any of it. They match the wording of a buyer's question to the wording of an answer. A transcript. A caption. A paragraph. Whoever wrote the best answer wins, whether they have 400 followers or 400,000.
AirOps puts this pretty plainly: one exceptional article on an under-followed page can out-cite a company page with ten times the audience and nothing worth citing on it.
The asset is the piece of content. Not the audience gathered around it.
Which makes this a quality game, not a quantity game
Here's my own read, from my own logs, layered on top of their data.
I see smaller creators with sharp, substantive content out-cite clickbait accounts with 10x the following. Constantly…and it makes complete sense. Clickbait is built for the click. It's vague on purpose. Substantive content answers the question. A creator who walks through a real implementation, names the tradeoffs, and shows the result has made something an engine can match to a real buyer question. "5 SHOCKING tools you NEED" matches nothing.
I can't emphasize this enough when vetting creators: do not go by raw follower count.
Vet the work. Does this person go deep in the field? Do they answer real questions with real specifics? Would their content hold up as a source in front of a skeptical buyer? That's what earns the citation. Reach is a bonus. Substance is the asset.
A personal receipt on this one. StackedGTM is about 21,000 people right now. I've been turned down by brands because my following was too small. Meanwhile I work with their competitors, and in those categories I'm one of the top cited creator sources on this exact subject, if not the top one. The brands that passed on my audience are now competing against my citations. This has now happened probably half a dozen times.
Two traps that will waste your budget
Before you move a dollar, two findings from the data will save you from spending it badly.
Trap one: the wrong engine. Google AI Overview pulls roughly one in every 9.5 citations from creator content, and one in 12 from YouTube alone. ChatGPT pulls 0.9% from creators total. YouTube inside ChatGPT is 0.1%. A rounding error.
Same video. Gold in one engine, invisible in the other. If your buyers ask their questions in ChatGPT and you fund a YouTube program because it worked for a brand whose buyers live in Google, you'll spend real money and get nothing back. Figure out where your buyers actually ask before you brief a single creator.
Trap two: the wrong platform for your industry. No two verticals lean on creators the same way. In B2B software, LinkedIn and YouTube carry the signal while Instagram and TikTok carry almost none. In consumer and retail, YouTube does most of the work. In fintech, comparison sites and financial publishers do the heavy lifting and creators add on top. In legal, practitioner Q&A and community threads shape answers more than polished professional content does.
Spreading budget evenly across platforms treats them all as the same opportunity. They aren't.
Where experts show up in the buying journey
One more cut worth knowing. Creator citations cluster at the edges.
Early, when buyers ask what something is and how it works, creators take about 7.7% of citations. Demos, guides, practitioner walkthroughs. Late, when buyers are close to a decision, creators come back at 4.7%. Firsthand proof that the thing works and fits.
In the middle, while buyers compare options, creators go quiet.
So don't brief creators to cover the whole journey. Brief them against the moments where engines already treat their work as proof: early education and final validation.
What I'd do heading into 2027
Start with the gaps, not the content. Pull your citation data and list the buyer questions where someone else's answer shows up instead of yours. Every question on that list is a brief waiting to be written. A lot of teams start with "what should we make." Wrong order. Start with "where are we losing."
Let the engine pick the platform. If your buyers ask in Google, YouTube is a weapon. If they ask in ChatGPT, it's a write-off. This one decision, made before any creative conversation, is the difference between a program that compounds and a budget that evaporates.
Recruit for scar tissue. The practitioner with fifteen years and 4,000 followers will out-cite the account with 400,000 and nothing to say. Read their work the way a skeptical buyer would. If it holds up, sign them. If you catch yourself impressed by the follower count, start over.
Brief the edges, skip the middle. Creators earn citations when buyers are learning and when they're deciding, and go quiet in between. One precise piece against one real question at those two moments beats ten generic ones spread across the journey.
Fund it like rent, not like a launch. Campaigns end. Citations don't. The brands that win this will be the ones paying experts to keep answering questions long after a campaign brief would have expired.
The last word
The engines have already voted. What experts show beats what brands say, and no amount of brand budget votes that back.
I sat on the other side of this for years, running growth inside some pretty sick brands, and I know exactly how the creator conversation dies internally. The attribution looks fuzzy. The budget feels soft. The deck loses to a paid channel with a cleaner dashboard. I get it. I killed so, so many of those proposals myself.
But while that debate runs its course, some practitioner in your category is quietly becoming the answer to your buyers' questions. I know because in a few categories, I'm that practitioner. The brands that fund the work get the citations. The brands that passed are competing against them.
You don't get to sit this shift out. You only get to choose whether the substance being cited in your category is yours.
Invest in (quality) creators.
All data from AirOps: 3.5 billion citations tracked August 2025 through June 2026, with engine, industry, and funnel analysis from a 93-million-citation snapshot across six engines. Brilliant work from Beca Ramón and the team. Go read the full report.