AI influencer discovery tools are changing how agencies cast creators

In use since January, the tool is built around an API deal with Meta, and suggests creators based on typical factors like subject matter and profile, as well as whether they’re participating in a given social media trend. 

Related Insights

The Creator Economy

A year after Unilever, the ad-funded creator economy is still catching up to its own ambition Read More

Shenda Loughnane, global brand president at Dentsu X, told Digiday the idea is to turn brands’ ability to “tap into culture” via influencers into a repeatable system.

For Dentsu, CATS is encoded at the heart of a “playbook” for creator marketing it’s recently established across its client roster that emphasizes the cultural capital held by influencers – and the use of paid spend across Meta’s Partnership and Reels Trending ad formats to create measurable impacts.

Skincare brands Galderma and Elizabeth Arden are among the clients currently using the approach; in the case of the latter, the agency observed a 14.3% increase in unaided ad recall among customers and a 41% rise in sales conversions from partnership ads.

“Brands really need to think hard about how they cut through the ‘sea of same’, and how they really try and win in an algorithmic environment,” said Loughnane.

Dentsu is far from the only agency relying on AI tools for creator discovery. Tech has become a key differentiator among indie and holdco creator businesses.

Related Insights

The Creator Economy

A year after Unilever, the ad-funded creator economy is still catching up to its own ambition Read More

Independent influencer agency Later uses an AI system that matches campaign briefs with candidate influencers and models potential performance of content on social media based on historic engagement data.

“That’s a much richer picture that gives me confidence that I’m going to get a high ROAS on my campaign,” said Scott Sutton, CEO of Later. He told Digiday the agency’s teams have been using AI tools for creator discovery for six months.

These applications of AI follow earlier tech interventions in the creator marketing workflow. First, marketers and agencies such as Goat, Obviously, Viral Nation and Influencer tackled campaign brief generation, and later moved to use AI tools to vet the creators they considered working with.

Now that discovery and selection tools have entered the picture, much of the modern creator marketing campaign process is becoming automated.

For Later’s Sutton, it’s a necessary response to a changing market. Brands now work with so many creators as they bid to beat social algorithms that manual campaign organization is unfeasible.

“More and more brands want to work with more creators,” he said. “The mechanics of operating creative programs in a highly effective way require you to use more influencers – generally, smaller to mid size influencers – in a more targeted way.”

Walmart, for example, now deploys “hundreds of thousands” of creators, according to Sarah Henry, the retail giant’s vp, head of content, influencer and commerce. It’s not picking influencers based on their follower count, Henry said during a speech last week at the ANA’s 2026 Media Conference, but rather on engagement metrics.

Using such a lens at scale wouldn’t be possible “without breaking our team, and the client,” said Kevin Blazaitis, president at Creo. “I would say we’re probably working with 30 or 40% more influencers on average per campaign.” 

The agencies using such tools maintain that their staff are able to exercise oversight over their recommendations; Blazaitis even suggested that Creo’s Discovery Agent could eliminate political or racial biases from the discovery process. 

“I don’t view this as a replacement for humans. It’s giving them a better starting point,” said Blazaitis.

Nevertheless their use means the judgment, experience and intuition once essential to creator selection are becoming reserved for select cases involving top tier talent.

For campaigns calling upon quasi-celebrity figures or household names, Sutton said, the white glove casting treatment was still necessary. But today’s armies of nano- or micro-influencers didn’t require such a close handling by humans.

“There are times when that depth is required,” he said. “Human plus AI is the best outcome. But these are large-scale data problems where AI is amazingly well suited.”