Google isn’t losing searches. It’s earning less from each one, and AI is part of the problem. OpenAI says ChatGPT now handles roughly one-fifth of the volume of search-like queries that Google does. They made that announcement just days after Google’s own earnings showed its AI push is changing how it makes money.
Adam Fry, who leads consumer product for ChatGPT, made the claim at an OpenAI dinner in New York on November 2. “ChatGPT’s volume of search-like queries is now roughly one-fifth the volume of queries of Google,” he said. He defined a search as “basically anything that a user would have traditionally gone to a traditional search engine for.” OpenAI also says ChatGPT now has 1.2 billion weekly active users, 250 million weekly voice users, and about 1.5 billion images generated a week.
However, OpenAI is private, so the figure hasn’t been independently verified. Fry named no Google baseline, time period, or region, and it isn’t clear whether several follow-up questions in one conversation count as one search or many. Outside estimates are lower. SEO firm Ahrefs estimated in February that ChatGPT handles about 12 percent of Google’s search volume for traditional queries, though a raw comparison of ChatGPT prompts to Google searches came to “nearly a fifth” because it counted coding and writing too. SEO platform seoClarity puts the share at 7.5 percent and calls it a modeled comparison, not a measured market share. Ahrefs also found Google sends about 190 times more traffic to websites than ChatGPT does.
That doesn’t mean Google isn’t under pressure, though. On Alphabet’s earnings call October 28, CEO Sundar Pichai said Search query growth was the highest it’s been in five years, driven by AI features. Queries in Google’s AI Mode run about three times longer than traditional searches, and one in five now uses voice, images, or documents. But revenue didn’t keep up. Google Search and other revenue grew 12 percent to $63.35 billion, down from 17 percent growth in the second quarter. Chief Business Officer Philipp Schindler said “monetization in AI Mode is lower,” and Google didn’t say how much lower.
In other words, AI is growing the number of questions people ask, and Google is earning less on the new ones. The market noticed. Alphabet’s stock fell from $324.89 on the day of the report to $283.82 the next day, and closed at $280.03 on November 3, 13.81 percent below its October 28 close. Alphabet also spent $51.355 billion on capital expenditures in the quarter, which pushed free cash flow to negative $9.065 billion, even as Google Cloud revenue grew 103 percent. To be clear: Google’s search volume isn’t shrinking, and its revenue is still growing. What’s shrinking is the pace of that growth, and the revenue each AI-powered answer earns.
The figures suggest the total pool of online questions is expanding rather than simply changing hands. The harder problem for both companies is turning conversational queries into money at the rate traditional search ads did. OpenAI is already working on its side. Fry said users don’t seem bothered by the ads now running in ChatGPT and that the company could be showing more. Anthropic, which plans to list on the Nasdaq with a roadshow expected around November 9, has said it intends to keep its Claude assistant free of advertising and lean on enterprise revenue instead. Its prospectus showed first-half 2026 revenue of $16.3 billion, according to AlphaSense. That contrast, ad-supported ChatGPT against ad-free Claude, shows how unsettled the business model for AI answers still is.
For anyone who depends on search traffic, from marketers to creators to publishers, the future of searches is important. Neither OpenAI nor Google breaks out conversational query volume or click-through rates in a way advertisers can benchmark. That’s why the industry has to rely on first-party data and testing rather than platform-reported share. And longer queries tend to carry more intent, so people are more likely to see specific use cases and comparisons over two-word keywords. And now there is a new goal for many: being cited inside an AI answer instead of just a Google search.