ChatGPT search volume: OpenAI says search-like queries in ChatGPT now run at roughly 20 percent of Google’s search volume. ChatGPT consumer product lead Adam Fry made the claim in a discussion covered by The Verge and analyzed by Search Engine Journal on November 3, 2026, drawing on the Big Technology conversation. It is a company figure — and it measures something different from website referral traffic.
For editors, SEO teams and product managers the distinction matters: query volume is not click volume. Copying the 20 percent line into strategy slides without that caveat mixes two metrics that currently sit far apart.
ChatGPT search volume: what OpenAI counts
Fry defines a search as any user question that triggers a web lookup inside ChatGPT. Pure definitions, rewrites or simple how-tos without a web call do not count. In the Big Technology conversation, SEJ quotes him: “Every time you ask a question and go search the web, that’s a search.”
What Fry did not disclose: which Google baseline OpenAI divides into, over which period, or which regions. The Verge notes that the source of Google’s volume remains unnamed. So “20 percent” is a relative vendor claim, not an independently audited market-share number.
Context from the same talk: Fry said weekly active ChatGPT users rose from one billion to 1.2 billion over the previous two months, according to SEJ and follow-up reports. That is also an OpenAI figure and sits outside the search definition — but it frames why the company is pushing the search narrative hard.
Referral data tells another story
Ahrefs tracks referral traffic across sites that use its free Web Analytics tool. SEJ cites the November 3 update: in October, across 118,671 sites, Google accounted for about 24.74 percent of tracked traffic and ChatGPT for 0.36 percent. That is not search-query market share; it is the share of visits those sources send to websites.
The gap is expected. ChatGPT answers many search-like questions inside the chat and sends far fewer outbound clicks than a classic SERP. Query volume can climb while referral share stays tiny. For publishers and affiliates, the click still pays the bills — not the prompt statistic.
Earlier Ahrefs modeling (February) had pegged ChatGPT search-equivalents near 12 percent of Google’s volume with a different method (prompt types rather than triggered web searches). The figures are not one-to-one comparable; they mainly show how strongly definitions move the result.
Google is growing at the same time — differently
Alphabet reports query growth as a rate, not an absolute count. SEJ points to CEO Sundar Pichai’s Q3 remark: year-over-year query growth is the highest in five years. Google says AI investments are lifting search overall — including AI Overviews and other generative features.
Neither Fry’s ratio nor Google’s growth rate proves that ChatGPT searches replace Google searches. Both can rise in parallel. Channel plans therefore need to treat chatbot search-like demand and classic organic search as separate lines with separate KPIs.
A sober reading
One-fifth of Google’s query volume would be striking — if the baseline holds — for a product that has only pushed search systematically for about a year and a half. It is still far from “replacing Google.” The Verge recalls Microsoft’s 2023 antitrust remark that Bing held about 3 percent share after more than a decade: a different metric, but a useful expectation check.
At code404zero we have already covered related OpenAI topics without the hype: LifeVal, the December merge of ChatGPT and Codex usage limits, and the raised bug-bounty rewards. Google’s school decision to keep Gemini and Search off together by default also shows how tightly search and AI are now coupled — regardless of OpenAI’s percentage.
Fry also talked about ads in the same setting. Per The Verge, he said ads had not durably hurt product usage and that OpenAI is testing richer sponsored product imagery rather than name-and-link only. That matters for search strategy: monetization incentives keep answers and recommendations inside the chat, which can keep referral shares low even as query volume grows.
What teams should do
If ChatGPT appears in a search strategy, keep two columns: (1) query/prompt volume with an explicit definition and (2) referral and conversion traffic. Put the 20 percent claim in column 1 with the note “OpenAI definition, Google baseline unclear.” Put Ahrefs-style referral data in column 2.
In practice that does not mean “search is dead.” It means answers increasingly form in chat while citations and clicks stay scarce. Brands need verifiable facts, primary sources and structured data — not only keyword rankings. Google still dominates most website traffic; referral numbers make that clearer than any prompt ratio.
What remains open
OpenAI has not published the Google baseline. It is also unclear what share of ChatGPT web searches runs through Bing, OpenAI’s own index or other sources — The Verge treats Bing and the in-house index as likely and Google routing as speculation. Whether ads and “Intelligent UI” change click behavior is not yet independently measured.
The takeaway stays factual: ChatGPT search volume is, by OpenAI’s count, roughly one-fifth of Google’s — measured as prompts that trigger a web lookup. Referral traffic in Ahrefs’ sample sits orders of magnitude lower. Plans should separate those definitions and cite primary reporting (Verge/SEJ on the Fry conversation, Ahrefs referral update, Alphabet earnings) rather than headlines alone.
