Takeaway: On Nov 3, 2026, ChatGPT consumer product lead Adam Fry said ChatGPT’s volume of search-like queries is now roughly one-fifth (about 20%) of Google’s query volume — a claim OpenAI shared in a fireside chat later published by Big Technology. The same briefing put ChatGPT at 1.2 billion weekly active users (up from 1 billion in about 60 days), 1.5 billion images generated per week, and about 250 million combined weekly voice/dictation users. For publishers and SEO teams, the headline number is attention-grabbing but definition-dependent: OpenAI counts each question that triggers a web search inside ChatGPT, not referral clicks to websites — and third-party referral trackers still show Google sending far more traffic than ChatGPT.
Primary reporting: Big Technology’s published interview with Adam Fry (Nov 3, 2026); secondary analysis from Search Engine Journal, The Verge, and 9to5Google.
What OpenAI actually claimed
Fry’s key line, as transcribed by Big Technology: ChatGPT’s volume of search-like queries is now “roughly one-fifth the volume of queries of Google.” He framed the growth as recent and steep — ChatGPT only gained reliable up-to-date web retrieval well after launch, when answers no longer stopped at a training cutoff.
In the same conversation OpenAI also highlighted product-scale metrics that matter for search competition:
- Weekly active users: about 1.0B → 1.2B in roughly 60 days (~100M users per month recently).
- Image generation: about 1.5 billion images created per week.
- Voice + dictation: about 250 million combined weekly active users speaking with or dictating to ChatGPT.
- Ads posture: Fry said ads have not hurt product usage and that OpenAI “could be doing this more,” per The Verge’s reporting on the same briefing — a monetization path that sits next to Google’s own AI Mode monetization pressure.
Separately in the same session, Fry discussed merging ChatGPT’s Chat and Work experiences — a product story we covered the same week in ChatGPT Chat and Work tabs merge. This article focuses on the search-volume claim and what it does (and does not) mean for SEO.
How OpenAI defines a “search”
Definition is the whole story. Asked whether a multi-turn conversation with one embedded search counts as a single query, Fry answered: “Every time you ask a question and go search the web, that’s a search.”
In practical terms, OpenAI’s search-like bucket is closer to “questions that would have gone to a traditional search engine” than to “every ChatGPT prompt.” Fry said early usage skewed to writing and coding help — tasks people often would not have typed into Google — while the last year saw more people, places, things, finance, and health questions that need live web grounding. Those are the intents ChatGPT increasingly routes through web search.
What Fry did not disclose, and every careful write-up notes:
- Which Google volume figure OpenAI used (daily queries, annualized total, which regions).
- The exact measurement window for the 20% comparison.
- Whether the Google denominator is worldwide or a subset.
- How ChatGPT’s unit (web-search-triggering questions) maps to Google’s classic query unit.
Treat “20% of Google” as an OpenAI-stated ratio of search-like activity, not an independently audited market-share statistic. Bing’s long-running ~3% search market share (from Google antitrust trial testimony) is a useful contrast for attention — ChatGPT is claiming query activity far above Bing’s classic share — but market share of clicks, ad dollars, and referral traffic remain different games.
Query volume ≠ referral traffic
This is the publisher-critical distinction. OpenAI’s 20% claim is about questions that trigger web search inside ChatGPT. Ahrefs’ referral tracker, cited by Search Engine Journal, measures something else: share of visits sent to websites. In October 2026, across about 118,671 sites using Ahrefs Web Analytics, Google accounted for roughly 24.74% of tracked traffic while ChatGPT accounted for about 0.36% (Nov 3 tracker update per SEJ).
| Measure | What it counts | What recent numbers say |
|---|---|---|
| OpenAI search-like queries | Each ChatGPT question that triggers a web search | ~20% of Google’s query volume (OpenAI claim; Google denominator undisclosed) |
| Ahrefs referral share (sample sites) | Visits attributed to Google vs ChatGPT | Google ~24.74% vs ChatGPT ~0.36% (Oct 2026 sample) |
| Alphabet Search query growth | Google’s own YoY query growth rate | Highest in five years in Q3 2026 (Pichai); not an absolute query count |
| ChatGPT → site clicks | Referral / citation traffic publishers can measure | Still a small fraction of Google referrals in public trackers |
Neither OpenAI’s ratio nor Google’s YoY growth rate proves ChatGPT is replacing Google queries one-for-one. Users can grow ChatGPT search-like usage while Google also posts strong query growth — especially if AI Mode and AI Overviews change how Google itself answers. For channel planning, always label which metric you mean: query activity, SERP impressions, or referral sessions.
Why the number still matters for SEO and GEO
Even with caveats, a credible claim that ChatGPT already handles search-like volume on the order of one-fifth of Google is a structural signal for search marketers:
- Habit shift is real. “Ask Chat” competing with “Google it” is no longer a joke slide — OpenAI is quantifying search-like demand inside its product.
- Citation and crawl access matter. If ChatGPT search answers cite the web, being fetchable by OpenAI’s search-oriented crawlers (and measuring AI referrals) is part of modern visibility — see our guides on robots.txt for AI crawlers and tracking AI referral traffic in GA4.
- Google is responding inside Search. Alphabet’s Q3 story already tied strong AI Mode query growth to lower monetization pressure on Search ads. OpenAI’s volume claim helps explain why Google keeps pushing AI Mode, AI Overviews, and related experiences.
- CTR and zero-click dynamics continue. Whether answers live in ChatGPT or in Google AI Overviews, publishers need frameworks for AI Overview CTR / halo effects and generative engine optimization (GEO) — classic blue-link SEO alone under-describes demand.
- Ads will follow attention. Fry’s comments that users largely tolerate ads, and that OpenAI could expand them, point to a second monetized answer layer alongside Google’s auction. Sponsored product imagery in ChatGPT is already expanding.
What publishers and SEO teams should do this week
- Split reporting columns. Track Google organic, Google AI surfaces (where measurable), ChatGPT/AI assistant referrals, and Direct separately — ChatGPT links often collapse into Direct when referrers are stripped.
- Audit AI crawler access. Allow search/citation bots you want visibility from; decide training bots separately (robots.txt guide).
- Prioritize cite-worthy pages. Clear answers, primary data, updated dates, and entity clarity help both Google’s generative surfaces and ChatGPT search citations — aligned with Google’s generative AI SEO guidance and GEO practice.
- Do not overreact to the 20% headline alone. Use it as a prioritization nudge, then validate with your own GA4, Search Console, and server-log evidence.
- Watch OpenAI’s ad and shopping surfaces. As ChatGPT monetizes answers, commercial SERP competition moves partially into chat UIs.
Context: Google is not standing still
Alphabet’s latest earnings narrative emphasized that Search query growth remains strong even as AI Mode changes monetization mix — covered in our Alphabet Q3 / AI Mode monetization analysis. OpenAI’s 20% claim and Google’s “highest query growth in five years” can coexist: total information-seeking grows, interfaces multiply, and click economics diverge by surface. The competitive question for 2026–2027 is less “did Google lose queries overnight?” and more “which answer surfaces own high-intent commercial demand — and which ones still send measurable visits?”
FAQ
Did OpenAI say ChatGPT has 20% of Google’s search market share?
No. Fry said ChatGPT’s volume of search-like queries is roughly one-fifth of Google’s query volume. That is an OpenAI activity comparison, not audited click share, ad share, or referral share. Market-share language in headlines often overstates what was measured.
What counts as a ChatGPT “search” in OpenAI’s number?
Per Fry: every time a user asks a question and ChatGPT goes to search the web. Multi-turn chats can contain multiple searches. Writing/coding prompts that never trigger web retrieval are outside that bucket.
Does ChatGPT already send 20% as much website traffic as Google?
Public referral trackers say no. Search Engine Journal cited Ahrefs Web Analytics showing Google at about 24.74% of tracked traffic versus ChatGPT at about 0.36% across ~118k sites in October 2026. Query activity inside ChatGPT and outbound clicks to publishers are different metrics.
How fast is ChatGPT’s user base growing?
In the same Nov 3 briefing, Fry said weekly active users rose from about 1 billion to 1.2 billion in roughly 60 days — on the order of 100 million users per month recently, according to Big Technology’s write-up.
What should SEO teams change because of this claim?
Add ChatGPT/AI assistant visibility to the measurement stack (crawlers, citations, GA4 AI channels), keep investing in cite-ready content, and avoid treating OpenAI’s 20% ratio as a reason to abandon Google SEO. Validate with your own traffic data.
Sources: Big Technology — Adam Fry interview (Nov 3, 2026); Search Engine Journal analysis of OpenAI’s search-volume claim; The Verge; 9to5Google; Business Insider. Internal context: track AI referrals in GA4, AI Overview CTR / halo, AI Mode monetization / Alphabet Q3, AI crawlers robots.txt, GEO guide, ChatGPT Chat/Work merge.