🔍 Read the full analysis: We Compared OpenAI’s ChatGPT Search Claim With The Available Data on ThorstenMeyerAI.com
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
OpenAI consumer product chief Adam Fry said ChatGPT handles roughly one-fifth as many search-like queries as Google, but OpenAI has not disclosed the comparison’s source, period or geographic scope. Independent modeled estimates are lower, while website traffic panels show Google sending far more visits than ChatGPT. The measures count different things, so they do not directly confirm or disprove OpenAI’s claim.
OpenAI’s Adam Fry said on November 2 that ChatGPT handles roughly one-fifth as many search-like queries as Google, but the company has not published the underlying calculation. A comparison with independent estimates, OpenAI’s usage data and website traffic panels finds no public basis for verifying the ratio—and shows that Google still sends substantially more traffic to websites.
Fry, who leads ChatGPT’s consumer product, made the statement during a conversation with journalist Alex Kantrowitz at an OpenAI-hosted dinner in New York. Big Technology published the conversation the next day; The Verge described the gathering as a press event. Asked what counts as a search, Fry said, “Every time you ask a question and go search the web, that’s a search.” The Verge reported that simple requests such as defining a word or explaining a basic how-to do not count under that description.
Fry did not identify which Google query figure OpenAI used, the time period or countries covered, or any supporting dataset. The Verge reported that he did not specify the source of Google’s search-volume number. Google had not commented on the claim in the reports cited by the source material. The statement itself is documented; its underlying ratio remains unverified outside OpenAI.
Two independent estimates put ChatGPT below the one-fifth figure, though both are models rather than direct measurements. Ahrefs estimated about 12 percent in February 2026, using an estimate of ChatGPT prompts and its own judgment about which prompts resemble Google searches. seoClarity estimated about 7.5 percent on October 28, combining projected ChatGPT message volume and selected message categories with an assumed rate of Google search growth. seoClarity called its result a modeled comparison, not a measured market-share figure.
Website referrals show a different gap because they count visits rather than queries. In Ahrefs’ October panel of 118,671 websites, Google accounted for 24.74 percent of visits and ChatGPT for 0.36 percent—about 69 times as much traffic from Google in that panel. StatCounter’s comparison of referrals to sites in its network put ChatGPT’s share of clicks from Google and ChatGPT between 0.08 and 0.41 percent per month from October 2025 through October 2026. Those figures do not measure how often people ask either service a question.
Infographic 01 · ChatGPT as a share of Google
Questions or clicks? ChatGPT is a fifth of Google, or less than 2 percent
OpenAI says one fifth. Independent models say less. Traffic panels say far less, because a chat answer often ends without a visit.
Counting questionsHow many search-like questions?
Counting clicksHow many visits reach websites?
All values: ChatGPT as a percentage of Google on the same yardstick. OpenAI’s figure is a company claim (striped bar); the two models are estimates; the panels see only their own sites. Bars scaled to 20 %.
Sources: Big Technology, 3 Nov 2026 · Ahrefs, 11 Feb 2026 · seoClarity, 28 Oct 2026 · Ahrefs traffic panel, Oct 2026 · StatCounter, Oct 2025–Oct 2026.
Powered by Thorsten Meyer AI
Search Use and Website Visits
The distinction matters to publishers and businesses deciding where to spend effort. A person may use ChatGPT to find information without clicking through to a website, while a search query does not necessarily result in a visit either. Query volume and referral traffic are different measures; neither set of figures can be substituted for the other.
The available traffic panels suggest that, for the websites and periods they cover, Google remains a much larger source of visits. That does not settle how many searches ChatGPT handles, but it does mean the one-fifth claim should not be read as evidence that ChatGPT is already delivering a comparable share of web traffic. Businesses may track whether they are cited in AI answers as well as conventional search referrals, while treating each as a separate metric.
For users, the reported figures also do not establish how accurate or complete ChatGPT’s answers are. When an answer affects a decision, opening its cited sources can help readers check the underlying information rather than relying on the answer alone.
Popular right now
Bose QuietComfort Headphones (2nd Gen), Wireless Over-Ear Noise-Cancelling Headphones, Black
Samsung 65-Inch The Frame Pro Neo QLED 4K TV (2026) Bundle with Modern Brown Bezel, HDMI Cables, and Home Theater Guide
De'Longhi Dinamica Plus Fully Automatic Espresso Machine
As an affiliate, we earn on qualifying purchases.
How the Estimates Differ
OpenAI’s comparison is described as a count of search-like questions that trigger a web lookup. The company did not provide enough detail in Fry’s remarks to reproduce the calculation. Other estimates use different definitions: Ahrefs classified a portion of estimated prompts as searches, while seoClarity selected “seeking information” and part of “practical guidance” in its model.
OpenAI’s published usage research offers another reference point, but not a direct search count. The source material says “seeking information” accounted for 20.2 percent of consumer ChatGPT messages in September 2026. That category is not the same as web searches, and the number alone cannot be used to verify Fry’s ratio. The source’s rough arithmetic suggests that reaching one-fifth of Google’s query volume would require about six in ten messages to trigger a web search, but that calculation depends on assumptions about the underlying volumes and definitions.
Fry also said ChatGPT had grown from one billion to 1.2 billion weekly users over the preceding two months. The source provides that as his statement, not an independently verified user count. Meanwhile, Alphabet said third-quarter Google query growth was its fastest in five years and search revenue rose 12 percent to $63.35 billion. Those company-reported figures indicate continued Google growth, but do not establish the size of ChatGPT’s search activity.
““A modeled comparison, not a measured market-share figure.””
— seoClarity
The Missing OpenAI Method
The main unresolved issue is how OpenAI arrived at the one-fifth comparison. Fry did not identify the Google benchmark, the dates or locations included, or the number of ChatGPT questions that triggered a web lookup. Without those details, outside researchers cannot reproduce the calculation or compare it precisely with the modeled estimates.
The independent figures also have limits. Ahrefs and seoClarity used assumptions to classify activity and estimate volumes; they are not direct counts of all searches. Traffic panels cover their participating websites and measure referrals, not total query volume. OpenAI’s “seeking information” category likewise does not show how many messages involved a web search. The available evidence therefore supports neither a direct confirmation nor a definitive rejection of Fry’s ratio.
What Better Disclosure Would Show
A clearer comparison would require OpenAI to publish its definition of a search-like query, the period and markets covered, the Google figure used and the calculation behind the ratio. Until then, researchers and businesses will continue to rely on estimates that are not directly comparable.
The next useful evidence would include updated, clearly defined usage data from OpenAI and further results from independent models and website panels. Alphabet’s reported search growth provides a measure of Google’s performance, but not a direct comparison with ChatGPT. For now, readers should treat Fry’s statement as an OpenAI claim without publicly disclosed supporting calculations, and keep query estimates separate from website-referral data.
Key Questions
What did OpenAI say about ChatGPT and Google?
Adam Fry said ChatGPT handles roughly one-fifth of Google’s query volume when counting search-like questions. He did not publicly specify the Google benchmark or provide the calculation.
Has the one-fifth figure been independently verified?
No. The statement is confirmed as something Fry said, but the underlying ratio has not been independently verified in the available material.
What do the independent estimates say?
Ahrefs estimated about 12 percent in February 2026, and seoClarity estimated about 7.5 percent on October 28, 2026. Both are modeled estimates with different definitions and assumptions, not direct measurements.
Does ChatGPT send as much website traffic as Google?
No, not in the cited panels. Ahrefs’ October 2026 panel recorded Google at 24.74 percent of visits and ChatGPT at 0.36 percent across 118,671 sites. Those results describe that panel, not every website, and measure visits rather than searches.
Why are query volume and website traffic different?
Query volume counts questions or searches, while referral traffic counts visits to websites. A ChatGPT answer may not lead to a click, so a high search-like query count would not automatically mean comparable website traffic.
Source: ThorstenMeyerAI.com
