ChatGPT cracks Google’s query monopoly
For two decades, no single competitor has taken a double-digit share of global search queries from Google. In July, ChatGPT did, while Google redesigned its search box for the first time in 25 years and the citation game splintered by platform. Here is what else moved the market.
Key takeaways
Ads in AI Mode don’t buy citations, earned beats paid.
AI Mode now serves ads on nearly 30% of commercial queries, but paying for one doesn’t improve a brand’s odds of being cited in the AI-generated answer itself. Visibility can’t be bought, it must be earned through PR, structured content and the sources each engine already trusts.
Named experts are outperforming company pages.
LinkedIn now supplies almost 1 in 8 social citations in AI search, up from 7.8% in January to 11.7% in May, and most of those citations go to individual people, not company pages. An untracked, inactive executive presence is now a visibility gap, not just a missed opportunity.
The GEO market is professionalising faster than most brands are executing.
92% of marketers say they plan to optimise for AI search, but only 40.6% are doing it. Tooling and agencies are racing ahead of real execution and the brands closing that gap now, engine by engine, have the advantage.
This month in AI search and visibility
Every month our team tracks what is moving in generative engine optimisation (GEO), AI search visibility and agentic AI, across the trade press, the platforms themselves and the conversations our clients are having. Here are the biggest stories of the moment: what happened, why it matters, and where to read more.
01: ChatGPT’s share of AI referrals hit 92.4%, and it’s still rising
For two decades, no single competitor has taken a double-digit share of global search queries away from Google. This month, ChatGPT did. New tracking puts ChatGPT at roughly 17% of all global digital search queries, with 900 million weekly active users processing 2.5 billion prompts a day. Separately, a Previsible study spanning 6.77 million AI-referred sessions across 166 sites over 19 months found ChatGPT now drives 92.4% of all standalone AI referral traffic, up from around 84% in December 2025, a share that is concentrating further even as more AI products launch.
The practical read for brands: “optimising for AI search” is, in practice, overwhelmingly optimising for one engine. Strategies built to spread evenly across ChatGPT, Gemini, Perplexity and Claude risk missing where the referral traffic and the buying attention actually is.
This is exactly the gap our AI visibility audits are built to close. Before we advise a client on GEO strategy, we map their brand’s presence engine by engine, not as one blended “AI search” score. Right now, most of those audits come back with the same headline: strong in the wrong engine, or invisible in the one that actually carries the traffic.
92.4% |
Share of all standalone AI referral traffic now driven by ChatGPT, up from ~84% in December 2025 |
FURTHER READING: Adapt Worldwide: AI search news roundup · Search Engine Land: SEO priorities in the age of AI search
02: AI citations are not driven by ad placements
AI Mode now serves a text ad on close to 29.45% of commercial queries but marketers tracking the same queries are finding that paying for the ad does nothing to improve the odds of being cited in the AI-generated answer itself. None of this should be surprising as organic search has run on the same principle for years. That gap between spend and citation is a defining data point: money buys proximity to the answer, not a place inside it.
The citation slots themselves are also getting harder to win, and different on every engine. ChatGPT’s citation pool is strikingly narrow, roughly 30 domains capture two-thirds of its citations on a given topic, and just under half of everything it cites traces back to Wikipedia (47.9%). Perplexity runs the opposite playbook, leaning on Reddit (46.7%) and rewarding very recent content, with pages under 30 days old earning 3.2x more citations. Google’s own properties (Business Profiles, Knowledge Panels, google.com itself) saw citations inside AI Mode surge 8.4x across the quarter, meaning brands are increasingly competing with Google itself, not just each other.
This is the exact case for earned-first thinking. Our PR citation strategy work starts from the assumption that no AI engine can be bought into, it has to be earned through coverage, structured content and the right domains, engine by engine. We’re already building citation plans this way for clients and the latest data just proved paid media can’t shortcut it.
29.45% |
Share of commercial AI Mode queries now carrying an ad with no measurable effect on citation odds |
FURTHER READING: B2the7: AI Mode ads, GA4 and AI traffic trends · Leapd: how ChatGPT, AI Overviews and Perplexity source information · BigGo: Google’s own domains surge in AI Mode citations
03: LinkedIn and personal expertise are becoming an AI citation channel
A new OtterlyAI study sampling citations found LinkedIn now accounts for nearly 1 in 8 social citations in AI search, with a 11.7% share in May, up from 7.8% in January. Perplexity and Google’s AI Overviews cite it heavily; Gemini barely does. The detail that matters most for brands here is that the majority of those citations go to individual people, not company pages.
That’s a meaningful shift in where credibility lives. A polished company blog post is now competing for attention with a named executive’s post on the same platform, and increasingly losing. Personal, named expertise is becoming a visibility channel in its own right, not a nice-to-have alongside brand content.
This is precisely why we run executive LinkedIn tracking for clients, not as a vanity metric, but because named-expert content is now out-citing branded content on one of AI’s fastest-growing source platforms. The brands with an active, tracked executive presence are already ahead of this shift; the ones without one are the company pages losing ground in the data above.
11.7% |
Share of AI social citations now going to LinkedIn, up from 7.8% in January 2026 |
FURTHER READING: OtterlyAI: LinkedIn AI search citations study
04: The GEO adoption gap is wider than the hype suggests
92% of marketers now say they plan to optimise for AI search, about as close to universal agreement as this industry gets. The catch is in the second number: only 40.6% are actually doing it today. That “everyone agrees, few are executing” gap is why GEO tooling (Otterly.ai, Rankscale, Ahrefs Brand Radar) is having a moment, and why agencies are formally repositioning around AI visibility rather than treating it as an SEO side-service.
The market is professionalising in real time, but unevenly: budget, tooling and public commitment are all racing ahead of the harder, slower work of actually restructuring content and coverage around how each engine cites.
40.6% |
Share of marketers actually executing a GEO strategy, versus 92% who say they plan to |
FURTHER READING: Omnibound: Generative Engine Optimization statistics · Optimist: SEO agencies pivot to AI visibility
05: Consumers trust AI search less, just as it gets fewer chances to earn it
AI search usage keeps climbing, but trust in it is heading the other way. The share of consumers who say AI-powered search is more helpful than traditional search has dropped from 82% a year ago to 54% now, a 28-point decline in twelve months. At the same time, changes tied to GPT-5.6 mean ChatGPT is asking around 90% fewer follow-up questions, so the first answer a user sees is increasingly the only one they get.
Put together, that is a shrinking margin for error at the exact moment users are growing more sceptical. Our thought leadership programmes are built to produce the sourced, expert-led content that AI engines (and increasingly sceptical users) treat as trustworthy on the first pass – let us know if you’d like to discuss how thought leadership can help make a difference.
54% |
Share of consumers who say AI search is more helpful than traditional search, down from 82% a year ago |
FURTHER READING: Search Engine Land: AI search adoption rises, consumer trust declines · DesignRush: ChatGPT traffic and AI search recap
Archetype SIGNALS
SIGNALS Archetype’s monthly brief on generative engine optimisation (GEO), AI search visibility and agentic AI. Every month our team tracks what is moving in generative engine optimisation (GEO), AI search visibility and agentic AI, across the trade press, the platforms themselves and the conversations our clients are having.
Archetype SIGNALS is delivered to clients who have subscribed to our newsletter. Sign up below to get the latest on AI search delivered straight to your inbox.
Sign up to our Newsletter
Sign up to our newsletter for monthly insights into the agency’s work, and the future of communications, marketing, media and AI.

