ChatGPT now handles search-like queries at roughly one-fifth of Google's search volume, according to OpenAI's own estimate. Adam Fry, the executive leading ChatGPT's consumer product, made the statement at a company-hosted event in New York on November 2. Big Technology published the interview on November 3. The underlying calculation has not been released for independent verification, and the figure is not a measured search market share.

In the original Big Technology interview, Fry described a search as a question that leads ChatGPT to retrieve up-to-date information from the web. Multiple web lookups within a longer conversation can count as separate searches. Writing and coding assistance, historically common uses of ChatGPT, do not automatically qualify under that definition.

That distinction matters for publishers and acquisition teams. Search events inside an AI assistant are not equivalent to outbound clicks to websites. The comparison does not establish that one-fifth of Google's users have moved to ChatGPT, that OpenAI owns 20% of the search market, or that its advertisers capture a comparable proportion of Google Search conversions.

The Verge noted several missing methodological details: OpenAI did not identify the Google query total used as its benchmark, the matching observation period or the geographic scope. Counting successive web-enabled questions in a conversation may also yield a different unit from the way traditional search engines describe queries and sessions.

Fry separately said ChatGPT had 1.2 billion weekly active users after gaining roughly 200 million in two months. Those vendor-reported figures cover the wider product. They are not the number of search users and provide no direct measure of publisher referrals, advertising impressions or sales.

For SEO teams, the useful response is to separate AI answer visibility from traffic and revenue measurement. Monitor when a brand is named or cited, then check actual referrals, on-site engagement and conversions in analytics. A rise in search-like questions does not necessarily translate to a comparable increase in website visits.

The number signals that ChatGPT is becoming an information-discovery destination. For media buyers, however, it is not a substitute for campaign-level evidence about inventory, CPC, CPA or return on spend. OpenAI's estimate should be cited as a company claim until its methodology can be assessed.

Sources

Big Technology — Adam Fry interview, published November 3

The Verge — independent reporting, November 3

Search Engine Land — measurement caveats, November 4

ChatGPT now handles search-like queries at roughly one-fifth of Google's search volume, according to OpenAI's own estimate. Adam Fry, the executive leading ChatGPT's consumer product, made the statement at a company-hosted event in New York on November 2. Big Technology published the interview on November 3. The underlying calculation has not been released for independent verification, and the figure is not a measured search market share.

In the original Big Technology interview, Fry described a search as a question that leads ChatGPT to retrieve up-to-date information from the web. Multiple web lookups within a longer conversation can count as separate searches. Writing and coding assistance, historically common uses of ChatGPT, do not automatically qualify under that definition.

That distinction matters for publishers and acquisition teams. Search events inside an AI assistant are not equivalent to outbound clicks to websites. The comparison does not establish that one-fifth of Google's users have moved to ChatGPT, that OpenAI owns 20% of the search market, or that its advertisers capture a comparable proportion of Google Search conversions.

The Verge noted several missing methodological details: OpenAI did not identify the Google query total used as its benchmark, the matching observation period or the geographic scope. Counting successive web-enabled questions in a conversation may also yield a different unit from the way traditional search engines describe queries and sessions.

Fry separately said ChatGPT had 1.2 billion weekly active users after gaining roughly 200 million in two months. Those vendor-reported figures cover the wider product. They are not the number of search users and provide no direct measure of publisher referrals, advertising impressions or sales.

For SEO teams, the useful response is to separate AI answer visibility from traffic and revenue measurement. Monitor when a brand is named or cited, then check actual referrals, on-site engagement and conversions in analytics. A rise in search-like questions does not necessarily translate to a comparable increase in website visits.

The number signals that ChatGPT is becoming an information-discovery destination. For media buyers, however, it is not a substitute for campaign-level evidence about inventory, CPC, CPA or return on spend. OpenAI's estimate should be cited as a company claim until its methodology can be assessed.

Sources

Big Technology — Adam Fry interview, published November 3

The Verge — independent reporting, November 3

Search Engine Land — measurement caveats, November 4

ChatGPT now handles search-like queries at roughly one-fifth of Google's search volume, according to OpenAI's own estimate. Adam Fry, the executive leading ChatGPT's consumer product, made the statement at a company-hosted event in New York on November 2. Big Technology published the interview on November 3. The underlying calculation has not been released for independent verification, and the figure is not a measured search market share.

In the original Big Technology interview, Fry described a search as a question that leads ChatGPT to retrieve up-to-date information from the web. Multiple web lookups within a longer conversation can count as separate searches. Writing and coding assistance, historically common uses of ChatGPT, do not automatically qualify under that definition.

That distinction matters for publishers and acquisition teams. Search events inside an AI assistant are not equivalent to outbound clicks to websites. The comparison does not establish that one-fifth of Google's users have moved to ChatGPT, that OpenAI owns 20% of the search market, or that its advertisers capture a comparable proportion of Google Search conversions.

The Verge noted several missing methodological details: OpenAI did not identify the Google query total used as its benchmark, the matching observation period or the geographic scope. Counting successive web-enabled questions in a conversation may also yield a different unit from the way traditional search engines describe queries and sessions.

Fry separately said ChatGPT had 1.2 billion weekly active users after gaining roughly 200 million in two months. Those vendor-reported figures cover the wider product. They are not the number of search users and provide no direct measure of publisher referrals, advertising impressions or sales.

For SEO teams, the useful response is to separate AI answer visibility from traffic and revenue measurement. Monitor when a brand is named or cited, then check actual referrals, on-site engagement and conversions in analytics. A rise in search-like questions does not necessarily translate to a comparable increase in website visits.

The number signals that ChatGPT is becoming an information-discovery destination. For media buyers, however, it is not a substitute for campaign-level evidence about inventory, CPC, CPA or return on spend. OpenAI's estimate should be cited as a company claim until its methodology can be assessed.

Sources

Big Technology — Adam Fry interview, published November 3

The Verge — independent reporting, November 3

Search Engine Land — measurement caveats, November 4

ChatGPT now handles search-like queries at roughly one-fifth of Google's search volume, according to OpenAI's own estimate. Adam Fry, the executive leading ChatGPT's consumer product, made the statement at a company-hosted event in New York on November 2. Big Technology published the interview on November 3. The underlying calculation has not been released for independent verification, and the figure is not a measured search market share.

In the original Big Technology interview, Fry described a search as a question that leads ChatGPT to retrieve up-to-date information from the web. Multiple web lookups within a longer conversation can count as separate searches. Writing and coding assistance, historically common uses of ChatGPT, do not automatically qualify under that definition.

That distinction matters for publishers and acquisition teams. Search events inside an AI assistant are not equivalent to outbound clicks to websites. The comparison does not establish that one-fifth of Google's users have moved to ChatGPT, that OpenAI owns 20% of the search market, or that its advertisers capture a comparable proportion of Google Search conversions.

The Verge noted several missing methodological details: OpenAI did not identify the Google query total used as its benchmark, the matching observation period or the geographic scope. Counting successive web-enabled questions in a conversation may also yield a different unit from the way traditional search engines describe queries and sessions.

Fry separately said ChatGPT had 1.2 billion weekly active users after gaining roughly 200 million in two months. Those vendor-reported figures cover the wider product. They are not the number of search users and provide no direct measure of publisher referrals, advertising impressions or sales.

For SEO teams, the useful response is to separate AI answer visibility from traffic and revenue measurement. Monitor when a brand is named or cited, then check actual referrals, on-site engagement and conversions in analytics. A rise in search-like questions does not necessarily translate to a comparable increase in website visits.

The number signals that ChatGPT is becoming an information-discovery destination. For media buyers, however, it is not a substitute for campaign-level evidence about inventory, CPC, CPA or return on spend. OpenAI's estimate should be cited as a company claim until its methodology can be assessed.

Sources

Big Technology — Adam Fry interview, published November 3

The Verge — independent reporting, November 3

Search Engine Land — measurement caveats, November 4