OpenAI says ChatGPT handles search-like queries at one-fifth of Google’s volume. The company did not disclose its comparison method or prove users switched.
ARTICLE BODY OpenAI says ChatGPT now handles search-like queries at about one-fifth of Google’s query volume. Adam Fry, the company’s consumer product lead, gave the estimate in a November 3 interview with Big Technology. He described it as OpenAI’s own comparison, but did not identify the Google baseline behind it. The headline may sound like a measure of people abandoning Google, but the figure counts search requests, not user migration.
What the 20% ChatGPT search figure measures
Fry described ChatGPT’s “search-like queries” as “roughly one-fifth the volume of queries of Google.” — Adam Fry, OpenAI consumer product lead. Fry clarified that each question triggering a ChatGPT web search counts as one query, even within a longer chat. He included current-information questions about people, places and things, plus some finance and health searches. He contrasted these with early uses such as writing or coding help.
That definition excludes many ChatGPT messages that do not trigger a web search. A request to edit a paragraph or write code would not count under Fry’s explanation. This counts searches, not users, conversations or clicked links. It tracks query events that invoke web search.
OpenAI’s search estimate lacks key details
The comparison comes from OpenAI, not an independent audit of the two companies’ systems. Big Technology’s interviewer asked how OpenAI compared ChatGPT’s search volume with Google’s, and Fry described the result. However, he did not disclose the Google figure used, the period covered or the geographic scope. The estimate is therefore difficult for outside researchers to reproduce. The interview gives a ratio but no raw query totals. It also does not show how many ChatGPT searches would otherwise have gone to Google. It compares the volumes each company counted, not tracked user switches.
Google said in May 2025 that it handled more than five trillion searches annually, citing internal data from January 2025. OpenAI did not say whether that figure formed the denominator for its one-fifth estimate. Nor did it explain if the two counts use the same rules for defining a query. Without those details, readers should treat 20% as a company-reported ratio, not a precise market-share calculation.
Search volume is not website traffic
Query counts and website visits measure different behaviour. Google users may click results. ChatGPT can answer without a website visit. A large number of AI queries would not automatically mean that publishers or businesses receive comparable referral traffic. One Google query can lead to several page visits. A ChatGPT search may end with a summary and no click. These different journeys affect referral traffic. A query-count comparison does not show how many visits each service sends publishers or how many ads users see.
Search Engine Journal cited an October Ahrefs tracker covering 118,671 websites. It found Google accounted for 24.74% of tracked traffic, compared with 0.36% for ChatGPT. That sample measures traffic sources for those sites, not the total number of searches on either service. It cannot confirm or disprove OpenAI’s query-volume comparison.
AI information searches do not prove Google dissatisfaction
Pew Research Center’s 2026 survey found that 42% of U.S. adults had used AI chatbots to search for information. The survey covered 5,119 adults in February and included chatbots such as ChatGPT, Gemini and Copilot. It measured reported chatbot search use, not frequency or Google switching.
Business Insider Africa’s headline suggests that users turned to ChatGPT because Google’s results had frustrated them. That may be the author’s interpretation, but neither Fry’s ratio nor Pew’s survey tests that explanation. The figures do not explain why users tried chatbots or whether they abandoned Google. Fry himself said many traditional searches now happen through ChatGPT when users want current information.
Google remains a separate benchmark
Business Insider Africa said Google had not immediately responded to its comment request. Google’s 2025 statement shows its search service remains enormous. But it cannot settle this comparison because the timeframes and counting rules may differ. The comparison also says nothing about which service provides better answers or earns more from search.
Ultimately, the estimate shows ChatGPT web searches have become significant. It does not show that Google has lost one-fifth of its users. The figure depends on OpenAI’s definition and an undisclosed Google baseline. Until the methodology is clearer, query volume should not be confused with traffic or satisfaction.
