When a consumer asks an AI assistant which shoe to buy, the answer may trace back to a YouTube video with a few thousand views, content produced by a creator a brand never briefed for that purpose. Influencer, a New York-based global creator marketing agency, believes brands have been generating those AI signals unintentionally through their existing campaigns for years. The firm’s new practice, Creator-First AEO, is designed to make the process deliberate.

Influencer introduced Creator-First AEO in August 2026 through a partnership with AI marketing platform Profound. The practice focuses on shaping the signals, content, and conversations that AI systems use when recommending brands. “The loop really is: Profound shows you how AI sees your brand. Then Influencer makes the creator content, and then Profound measures the change,” says Caspar Lee, co-founder of Influencer. “That missing piece has always been really difficult for people, which is actually making the content that’s going to create an impact to what you’re measuring.”

Influencer was founded in 2015 by Caspar and Ben Jeffries. Caspar began posting on YouTube in 2010, later collaborating with artists and celebrities including Ed Sheeran and Dwayne Johnson, and brought that firsthand understanding of platform dynamics to the agency’s founding premise. The firm now operates with more than 200 staff across North America, Europe, and the Middle East, and reports more than 6,000 campaigns delivered for brands including Spotify, Google, Nike, and Coca-Cola. 

Caspar’s argument is that the influencer campaigns the agency has run for a decade were already feeding AI recommendation systems before anyone developed infrastructure to track it.

Every Influencer Campaign Has Been Building AI Signals. Most Brands Missed It.

Profound’s data has surfaced findings that challenge standard assumptions about what drives AI recommendations. View count is not the primary factor. Caspar describes reviewing data for an unnamed shoe brand and finding that a YouTube video with 35,000 views accounted for a “massive percentage” of why AI systems recommended that brand over its rivals, while more-viewed videos contributed less.

The determining variable, he says, is structural clarity. Whether a video’s title and description allow a large language model to identify what the video actually demonstrates determines whether that content registers as evidence inside AI systems. “If that’s not properly labeled and the title of that video and the description of that video don’t easily identify what is actually going on in the video, the LLMs aren’t necessarily picking up as well,” Caspar explains.

The implication is that content optimized for human attention and content that functions as AI evidence are not the same thing. “You really see how some of the videos doing so well are just the ones that are really well titled and well formatted versus the ones that are really the biggest, most viewed videos,” he says. “When you can combine that with the biggest, most viewed videos, you have a really powerful opportunity.”

The data variability extends beyond a single brand. AI sources vary significantly by product category, including within a single brand’s portfolio, and can shift week to week. “You could look at one particular vertical, and maybe YouTube is a huge contributor to what it’s saying to consumers about that vertical. And then you look at another vertical that you might assume would be the same, and it’s completely different,” Caspar says. “Our clients are already doing loads of content for the social platforms. They now just need to optimize it.”

Profound Marks the Homework Influencer Does

When a brand joins Creator-First AEO, Influencer brings them into the Profound platform to identify where AI representations of the brand are incomplete or inaccurate. From those findings, the agency builds a creator strategy targeting the gaps, deploys content, and re-measures on Profound to assess whether the AI’s treatment of the brand has shifted. The cycle repeats.

“We work with our clients in a very high-level, bespoke way,” Caspar says. “They work with us, and we make sure that they’re using Profound’s data to the best of its abilities and then using our platforms to make sure the content creation is done to the best of its abilities.”

Influencer chose to partner with Profound rather than develop AEO measurement capabilities internally, citing Profound’s position in the category and the logic of specialization. “It’s similar to how Salesforce is now working with Anthropic,” Caspar says. “Salesforce doesn’t have to build the leading LLM. They can work with the leading one.” Influencer’s existing technology infrastructure includes first-party API access across TikTok, YouTube, Pinterest, Snap, Meta, and Twitch; building a competing AEO measurement platform would require a different kind of investment in a different kind of capability.

The arrangement also addresses a credibility question: Profound measures outcomes that Influencer’s campaigns are intended to produce, meaning measurement and execution sit with two separate companies. “The numbers will be the numbers, good or bad,” Caspar says. “Sometimes, we’ll say, ‘Do this.’ And it might not work that week. And that’s fine because this is something you have to constantly improve on and work on.”

When AI Joins the Audience, a Small Creator Can Outperform a Big One

Creator-First AEO shifts which creators brands should prioritize. Caspar argues the dynamic mirrors a pattern already present in Influencer Marketing, where high-reach talent does not always drive outcomes, and suggests AEO will make that divergence more pronounced.

“You could have a big-name influencer who gets a million views but can’t convert at all, because people don’t go to them for that kind of thing,” he says. “This is going to be even more exaggerated.”

A creator with a small but authoritative audience in a specific product category, whose content AI systems regularly cite when users ask for recommendations in that category, may deliver more measurable brand value than a creator with a generalist following ten times the size, according to Caspar. He frames this as an expansion of creator value to brands rather than a new constraint. “AEO is picking up the content creators are already doing right now,” he says. “We’re just optimizing it and proving its worth. What that’s going to do is just bring even more people into the Influencer Marketing category.”

The Line Between Earned Authority and Manufactured Signals

Any practice designed to shape what AI says about a brand raises the question of where optimization becomes manipulation. 

Caspar draws the line through factual accuracy and genuine advocacy. “You need to continue to be authentic about the collaborations you take on,” he says. “If you just start taking on every brand, you’re not going to become an authority anymore.”

He extends the point to brands directly. Favorable content cannot correct a genuinely negative product reality. “If there’s something truthful that’s truly negative about your brand, you’re not going to be able to fix that by putting out a good few stories,” he says. “AI is very good at picking that up, too.”

Caspar draws on the SEO analogy for the competitive dynamics. Tactics that artificially inflate a brand’s AI representation, such as fabricated forum posts or fake articles, worked temporarily in early search optimization before platforms adapted. “There was a time maybe in SEO where that worked, but it stopped working,” he says. The better posture, in his framing, is to put truthful information forward through authoritative voices rather than manufacture it. The risk of doing neither is that AI ends up with an incomplete or inaccurate picture of a brand by default.

Creator Content Follows Consumers Into AI

The framing question Caspar returns to is whether optimizing for AI systems sits in tension with Influencer’s stated mission of humanizing brands. His answer collapses the distinction. Creator content that improves a brand’s AI representation ultimately reaches human consumers through the recommendations they receive when making decisions. “Ultimately, this content is going to help people make decisions and purchase decisions,” he says. “That’s the business we’ve been in for the last 10 years.”

Around three weeks in, Creator-First AEO has no published results, and Influencer is not the first agency in the space, as martech company Later launched a creator AEO offering earlier this year. “We’ve never claimed to be the very first people to ever do AEO,” Caspar says, “but we are definitely pushing it forward.”

For Caspar, Creator-First AEO is an extension of a shift that has been accumulating since creator marketing emerged: the movement of advertising spend and influence away from broadcast channels toward individual voices. AI represents one more surface where that shift is playing out. 

“Those same people are now going to impact another thing that people are using,” he says. “That can only be good, because it means your content is more valuable. And it means that more creators will be able to fund their content, which ultimately will be viewed by people.”

Subscribe to Our Newsletter

Check Out Our Podcast