ChatGPT Search Volume at Roughly 20% of Google Search, OpenAI Says

By Amara Osei

ChatGPT’s search activity is “roughly one-fifth (20%) of the volume of queries of Google,” OpenAI consumer product lead Adam Fry said at an OpenAI press event and chat held the day before the source report.

How large is ChatGPT’s search footprint?

OpenAI cited the estimate after pushing ChatGPT into AI-powered search, meaning search functions that use AI to retrieve or summarize information, about two years ago. The report describes Google Search as the dominant service and Microsoft’s Bing as a distant competitor with a single-digit share “at any given time.” It also calls ChatGPT the biggest rival Google Search has had.

The 20% estimate needs a caveat about method. The report does not identify the source of Google’s search-volume data or disclose a matching method for the two services. It gives no absolute query counts, comparison period or geography. The figure does not establish a settled search-market share.

OpenAI’s assertion may nonetheless change how agencies compare Google Search with ChatGPT. If directionally accurate, it shows that web search inside an AI assistant can approach the query volume of a long-established search engine. The report also links that scale to Google’s extensive expansion of AI in Search.

Takeaway: ChatGPT is a large search competitor on OpenAI’s claim, but the evidence does not establish a 20% search-market share.

What counts as a “search”?

Fry defined OpenAI’s unit of measurement: “Every time you ask a question and go search the web, that’s a search.”

Under that definition, the measure captures a user asking a prompt and then going to the web, not every ChatGPT interaction. The comparison depends on how OpenAI classifies those actions and on the Google denominator it uses. It is not a direct comparison of users, pages, clicks or ad views.

The report does not set out a collection method, sample or absolute volume for Google queries. It also does not explain how Google’s comparable query count is defined. The estimate does not establish:

• 20% of Google Search users use ChatGPT.
• ChatGPT captured 20% of overall search traffic.
• Google lost 20% of its search-advertising market.

Takeaway: The metric captures web-search use initiated from ChatGPT, not the full scale of either company’s search business.

How could search growth be monetized?

Google has acknowledged “lower monetization,” meaning weaker conversion of product use into revenue, as it expands AI in Search, according to the report. The source also suggests OpenAI faces a comparable challenge, though it supplies no revenue data to test that point.

The Verge reported that Fry said “most people do not seem to mind ads” in ChatGPT. He added: “we could be doing this more, actually.” His wording points to room for a larger advertising role. It does not disclose ad formats, pricing, campaign tools, a launch schedule or revenue.

Fry’s claim about user sentiment has no survey method, sample size or response data in the source. The statement records OpenAI’s view of user tolerance; it does not quantify that tolerance or demonstrate advertising acceptance at scale.

Takeaway: OpenAI sees room for more advertising, but the source provides no evidence of advertising scale or revenue.

Does personalization strengthen the case?

Fry also said “people really love” ChatGPT’s personalization and understanding. Personalization means adapting responses to a user’s preferences and context. The report does not explain which ChatGPT features Fry included or how OpenAI measured the reaction.

The statement gives OpenAI a second argument beyond query volume. Search activity shows demand for information, while personalization and understanding are presented as reasons to return to the assistant. But the report offers no data on whether users return, how often they use ChatGPT or which preferences they value.

The omission matters for advertising. An assistant that understands a user may support relevant messages, but the report does not show that personalization improves ad performance, ad acceptance or advertiser outcomes. Any link between understanding and commercial results remains an inference.

Takeaway: Users may value ChatGPT’s personalization, but OpenAI has not quantified that value or its commercial effect in the source.

What should marketers take from the number?

For agencies and advertisers, the estimate puts ChatGPT in the same planning conversation as Google Search. The report does not compare query intent, meaning the purpose behind a search, audience quality, conversion, available ad placements or revenue between the platforms. The 20% number cannot support a claim that Google has lost one-fifth of its advertising market or that OpenAI commands an equivalent share.

Google reports lower monetization from its AI Search push; Fry says most people do not seem to mind ChatGPT ads and that OpenAI could add more. The report frames both companies as facing monetization questions, though it quantifies neither.

Takeaway: If OpenAI can expand ads while preserving the personalization users reportedly value, ChatGPT’s search scale could become revenue; the source does not yet show that outcome.