OpenAI Just Admitted It: ChatGPT Search Volume Has Reached 20% of Google's—Time to Move the Budget
Business Insider published a report early on November 4. Adam Fry, OpenAI's consumer product lead, confirmed a number on the record: search queries on ChatGPT now amount to roughly 20% of Google Search's volume.
That figure came from OpenAI itself, not a third-party estimate. When a company volunteers "we've reached a fifth of the giant's volume," it usually means one of two things—the data is strong enough to show off, or the fundraising story needs it. Either way, the 20% magnitude speaks for itself.
20%
Google processes billions of searches daily. Twenty percent of that is a height no single competitor has ever touched—the last company that made Google nervous was TikTok, and TikTok took watch time, not search itself.
User behavior changed a while ago. Ad budgets haven't.
A Pew Research survey from this summer fills in the fuller picture: 42% of US adults have used AI to search for information, making it the number one AI use case—ahead of entertainment, news, and medical advice. Meanwhile, 38% of working adults use AI for work tasks.
By product: 44% of US adults have used ChatGPT. Second place, Gemini, sits at 24%. A clear gap.
Connect the numbers: nearly half the population already uses AI to find information, most of them through ChatGPT, and ChatGPT's search volume has reached 20% of Google's. The user migration is mostly done. Marketing budgets remain parked in old channels. That lag is the opening.
AI recommendations are rewriting purchase decisions
If search migration is a traffic-level story, Adobe's survey released November 3 moves the battlefield to the wallet: 59% of shoppers say they would switch brands based on AI recommendations.
This number means very different things for big and small brands. For incumbents, decades of accumulated mindshare no longer form an absolute moat in front of an AI answer—AI doesn't recommend you because your ad spend is large. For smaller brands, it opens an advancement channel that never existed before: if your product information is clearer, more credible, and easier to cite, you can intercept the buyer at the moment of recommendation.
The scales are tipping toward information quality, not budget size.
Pichai's line from last year, revisited
When Google CEO Sundar Pichai addressed AI competition last year, he said chatbots had reached hundreds of millions of users but Google's search volume was still growing: "It's far from a zero-sum game. TikTok came along, everyone used TikTok, and YouTube grew during that period."
Fair enough at the time. But the data now adds a footnote: Google's search volume may still be rising, yet ChatGPT's is rising faster—fast enough to have taken 20%. The pie is growing, sure. The changing cut of it is what matters.
There's a more practical layer: the act of searching itself is changing. A user who Googles "best running shoes" gets ten blue links and does the comparing. A user who asks ChatGPT gets one direct answer with a handful of cited sources. The first game is about ranking. The second is about who gets written into the answer. Entirely different games.
Three concrete moves for brands
- Move GEO from "experimental budget" to "fixed budget." Twenty percent of search volume isn't a forecast—it's an accomplished fact. Brands still "trying it out" are watching the window narrow.
- Audit your product information's readability for AI. 59% willingness to switch brands means AI recommendations move sales directly. Are specs, features, and use cases consistent across your site, retail listings, and third-party reviews? AI citing conflicting information is worse than citing none.
- Start logging AI channel data now. Citations, mention positions, referral traffic, conversions—track them separately. In three months you'll know whether to double down or adjust.
Twenty percent is a milestone, but it reads more like a mid-race split. User behavior has already crossed the river. Brands still watching from the bank are paying a rising cost for the hesitation.