The Brands Winning AI Search Aren’t The Biggest. They’re The Most-Cited.
As the temperature in London climbs, you’re nearly ready to press “buy” on flights to the South of France, chasing a swim and the sea. But planning the trip this time looked nothing like it used to. Instead of scouring Skyscanner for flights, asking friends for recommendations and trawling Reddit for tips, you planned the whole thing with Claude and ChatGPT.
And your behaviour is becoming the norm. According to McKinsey, 50% of consumers already use AI-powered search today. Indeed, per ZS, “Forrester’s 2025 Buyers’ Journey Survey found that generative AI is now the single most cited interaction type for purchase research – ahead of vendor websites, peer recommendations and analyst reports.”
When ChatGPT, Gemini, Perplexity or Google AI Overviews answer a question like “where should I stay in Nice,” they don’t hand back a list of links. They name suggested brands and explain why. This behaviour is evolving across every category, from flights to France to ingredients for dinner tonight.
The Data Shows This Is Where Influence Actually Happens Now
AI models cite a brand’s own domain directly in only about 1 in 6 responses. Instead they lean on third parties – typically credible publishers, review platforms, Wikipedia, Reddit. In categories like CPG and financial services especially, publisher and user-generated content make up more than 65% of the sources that AI search cites.
Additional research backs this up:
- 62% of AI brand citations trace back to earned and third-party signals
- Third-party voices carry roughly 3x the weight of anything a brand says about itself
- ChatGPT alone weights editorial coverage at 30–40% of what shapes a citation
Brands need to shift how they think about influence. It’s no longer enough to optimise owned content for visibility in search. Brands now need to consider how they’re being spoken about and featured by third parties, because that’s what these models are actually learning from.
The important thing isn’t just where AI looks – it’s where a brand can actually move the needle: third-party content is the lever, far more than the brand’s own site.
What Winning Looks Like
These AI models learn from what credible publishers put in front of them. Content that’s genuinely informative, written in the expert voice of the publisher it lives on, teaches a model something a brand’s own page never can. It shows that other credible voices vouch for your brand, consistently, in the same language.
It’s not one ranking to win. It’s thousands of narrow territories. Nobody wins “places to travel.” Someone wins “the best adventure spots with water near Sydney.” One article in one of those territories doesn’t do much. But the same claims, repeated across many credible publishers, does.
The Opportunity
The simplest way to think about it: this is a first-mover advantage. For challenger brands, it’s a shot at territories the market leaders haven’t claimed yet. For B2B and technical categories, it’s becoming the cited authority on the specific, narrow questions buyers are actually asking. It’s also just better traffic: visitors who arrive via AI search convert at roughly 4.4x the rate of traditional organic search visitors.
A brand doesn’t need to build that authority for AI search from scratch. It already exists – it just belongs to someone else: the publishers who already cover that category credibly and consistently. Winning it takes working with a range of those publishers to build out a strategy, and getting a number of them to mention the same brand, in the same terms, consistently enough that it reads as a pattern rather than one lucky placement. One article isn’t going to move the needle. The signal that moves a model is the same claims, names and category language showing up across many credible outlets, not one.
And it isn’t a one-time fix: per Ahrefs, AI Overview results change for the same query roughly 70% of the time between observations, with nearly half of citations swapped out entirely on each refresh. This is ongoing work, not a placement you make once and leave alone. It also isn’t instant – citations build over weeks. The honest early signal is content going live and getting indexed, not an overnight spike in mentions.
Why We’re In This Conversation
Direct media partnerships have always been operationally hard – every publisher runs its own briefing process, its own formats, and that multiplies fast across a network.
That’s the problem Avid was built to solve: one brief, into many publishers, without the manual chasing, version control and back and forth within a campaign that normally comes with it. Successful GEO needs exactly this. A consistent, credible presence across many publishers is what earns citations. Coordinating that at scale, without it collapsing into fifty separate conversations, is the hard part. It always has been.
We didn’t arrive at GEO from an SEO or PR background. We arrived at it because we already run this exact content and media partnerships mechanism, every day, and have done for hundreds of partnerships. GEO didn’t invent that need. It just raised the stakes on solving it properly.
Georgia Faure, Head of Market Activation & AI Projects | Avid Collective
Comments (0)