AIThis post was created with the assistance of artificial intelligence (AI).

🔍 Read the full analysis: A Closer Look At OpenAI’s Claim That ChatGPT Handles 20% Of Google Searches on ThorstenMeyerAI.com

Before you orderOffer from Amazon

Get the latest gadgets delivered free with Prime

  • Fast, free delivery on millions of items
  • Prime Video, Amazon Music and more included
  • Member-only deals all year
Start your free Prime trial Free trial for eligible customers · Cancel anytime
As an affiliate, we earn on qualifying purchases.

TL;DR

OpenAI consumer product head Adam Fry said ChatGPT handles roughly one-fifth as many search-like queries as Google. The statement is confirmed, but OpenAI has not disclosed the comparison period, geography, Google benchmark or underlying data. Independent estimates are lower and website traffic panels show a much larger gap in visits sent to sites.

OpenAI’s Adam Fry said on November 2 that ChatGPT handles roughly one-fifth as many search-like queries as Google, a comparison that could suggest the chatbot is becoming a substantial search alternative. But OpenAI has not disclosed the underlying data or exactly what it counted, and independent estimates and website traffic measurements produce sharply different results.

Fry, who leads ChatGPT’s consumer product, made the statement during a fireside chat with journalist Alex Kantrowitz at an OpenAI-hosted dinner in New York. Kantrowitz published a transcript in his Big Technology newsletter on November 3. The Verge, which described the gathering as a press event, also reported on the remarks. This was a spoken claim, not a published study or formal company report.

Fry described a search as each occasion when a user asks a question and ChatGPT goes online to search the web. The Verge reported that the definition excludes some exchanges, such as asking for a word definition or a simple how-to explanation. Fry also said ChatGPT had grown from one billion to 1.2 billion weekly users over two months. The source material does not specify the measurement behind that user figure.

OpenAI did not identify which Google query count it used, the period or countries covered, or the data supporting the ratio. Google had not commented on the claim in the reporting cited by ThorstenMeyerAI.com. The claim is verified as something Fry said; its accuracy cannot be independently verified from the information released.

At a glance
analysisWhen: Claim made November 2, 2026; reported N…
The developmentAdam Fry said ChatGPT’s search-like query volume is roughly one-fifth of Google’s, prompting a comparison with independent estimates and traffic data.
Thorsten Meyer AI
Five yardsticks

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?

  • OpenAI (Adam Fry)search-like questions that trigger a web lookup · Google baseline not named≈ 20 %
  • Ahrefs modelprompts of a kind people used to search on Google≈ 12 %
  • seoClarity modelseeking information + a quarter of practical guidance≈ 7.5 %

Counting clicksHow many visits reach websites?

  • Ahrefs traffic panelvisits sent to 118,671 websites, Oct 2026≈ 1.5 %
  • StatCounterclicks to sites in its network, monthly range0.08–0.41 %

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 Is Not Website Traffic

The comparison matters because the phrase “search volume” can refer to different behaviors. A person may use ChatGPT to get an answer without clicking through to a website. That can make conversational query counts look substantial while producing far fewer visits for publishers, retailers and other site operators than conventional search.

In Ahrefs’ panel of 118,671 websites, Google accounted for 24.74% of visits in October 2026, compared with 0.36% from ChatGPT. That is a measure of traffic reaching those sites, not a count of searches. StatCounter’s comparison of clicks from search engines or chatbots to sites in its network found ChatGPT’s share of Google-versus-ChatGPT clicks ranged from 0.08% to 0.41% per month between October 2025 and October 2026. These panels do not settle how many search-like prompts ChatGPT receives, but they show that query comparisons should not be treated as equivalent to referral traffic.

For businesses and publishers, the distinction affects how they track audience reach and investment. ChatGPT citations may become relevant to visibility, but the cited traffic data does not show that chatbot referrals have replaced visits from Google. For users, an answer that includes sources still depends on the quality of those sources; following a citation can help verify information when accuracy matters.

Products Worth Considering

As an affiliate, we earn on qualifying purchases.

How the Estimates Differ

Two independent estimates cited by ThorstenMeyerAI.com place ChatGPT’s search-like activity below Fry’s one-fifth figure, but both are models rather than direct measurements. Ahrefs estimated about 12% of Google’s search volume in February 2026. Its Patrick Stox started with an estimate of 2.5 billion ChatGPT prompts per day from OpenAI’s 2025 usage study and judged that 65% were tasks people might otherwise take to Google. Stox cautioned that the proportion depended on how “search” was defined.

On October 28, seoClarity estimated a ratio of about 7.5%. It extrapolated ChatGPT message volume using OpenAI user-growth figures, counted all “seeking information” messages and a quarter of “practical guidance” messages as searches, and compared the result with an estimate of five trillion Google searches per year plus assumed 20% growth. The company called this a modeled comparison, not a measured market-share figure. Its assumptions and category choices differ from Ahrefs’ method and from Fry’s stated definition.

OpenAI’s own published usage research offers another point of comparison: “seeking information” accounted for 20.2% of consumer ChatGPT messages in September 2026. That category is not necessarily the same as web searches, since not every information-seeking message must trigger a web lookup. ThorstenMeyerAI.com argues that reaching one-fifth of Google’s volume would, by its rough arithmetic, require about six in ten total ChatGPT messages to trigger a web search. The underlying definitions and periods are not aligned well enough to treat that calculation as a definitive test.

“ChatGPT’s volume of search-like queries is now roughly one-fifth the volume of queries of Google.”

— Adam Fry, head of ChatGPT’s consumer product, as quoted in Big Technology

The Missing Comparison Details

The central unknown is what OpenAI divided by what. Fry did not name the Google query-volume source, reporting period, geographic coverage or underlying ChatGPT counts. Without those details, outside researchers cannot reproduce the one-fifth estimate or determine whether it includes only searches that trigger web browsing or other search-like interactions.

The available figures also cover different periods and behaviors. OpenAI’s September message categories, Ahrefs’ February model, seoClarity’s October model and the traffic panels’ site visits are not directly comparable. The traffic data records referrals to sampled websites, not chatbot prompts, while the models make assumptions about what users would otherwise search. No source cited in the report provides a single independently audited count covering the same definition, geography and time window for both services.

What More Data Could Show

A clearer assessment would require OpenAI to publish the definition, time period, geographic scope and data behind its ratio, alongside a comparable Google figure. Until then, the one-fifth number should be treated as an attributed company claim rather than an independently established measure of search-market share.

Businesses can track chatbot referrals and citations alongside conventional search traffic, while keeping those measures separate. Future estimates may change as services and measurement methods develop, but the source material does not identify a scheduled release of further data from OpenAI or Google. The next meaningful test will be whether either company publishes comparable figures that outside analysts can examine.

Key Questions

Did OpenAI confirm that ChatGPT has one-fifth of Google’s search volume?

Adam Fry said ChatGPT’s search-like query volume was roughly one-fifth of Google’s. OpenAI has not provided the data and definitions needed for independent verification, so the statement is confirmed as a claim, not as an independently established fact.

Fry said each time a user asks a question and ChatGPT searches the web counts as a search. The Verge reported that some exchanges, such as word definitions and simple how-to explanations, do not count under the description given.

What do independent estimates say?

Ahrefs estimated about 12% of Google’s search volume in February 2026, while seoClarity estimated about 7.5% on October 28, 2026. Both are modeled estimates based on assumptions, not direct measurements of a shared definition.

Does one-fifth of Google’s search volume mean ChatGPT sends comparable traffic to websites?

No. Query volume and clicks to websites are different measures. Ahrefs’ October panel recorded Google as the source of 24.74% of visits and ChatGPT as the source of 0.36% across 118,671 websites.

What information is still missing?

OpenAI has not specified the Google benchmark, time period, countries covered or supporting data for Fry’s comparison. The cited figures also use differing definitions, so a like-for-like independent comparison is not yet possible.

Source: ThorstenMeyerAI.com

You May Also Like

Deepfakes, Disinformation And Data: The First AI Midterms

AI-generated campaign ads, foreign deepfakes, chatbot use and a Georgia data-access incident are shaping the final week before the midterms.

Software Engineering At Google & CS Education With Titus Winters

A Pragmatic Engineer podcast revisits Titus Winters’ Google engineering work and his arguments for changing computer science education.

AI Enthusiasts Show Deep Devotion To Claude With Unusual Tribute At ‘Funeral’ Event

Anthropic employees reportedly held a funeral for a retired Claude AI model, highlighting emotional attachment to the chatbot, though details remain unclear.

Anthropic’s Claude Research Offers A New Angle On Software Proofs

Anthropic and Theorem report machine-checked proofs for 239 fuzz-test entry points, while warning the results cover a narrow software configuration.