ChatGPT Search Reaches About One-Fifth of Google’s Query Volume

ChatGPT’s search activity has reached roughly one-fifth of Google Search’s query volume, according to OpenAI. That’s a substantial comparison, but it describes reported activity rather than market share: it doesn’t tell us how many people use each service, how often they return, or whether they use ChatGPT instead of Google. The figure suggests the chatbot is becoming a sizable search alternative, though the underlying comparison leaves room for interpretation.

At an OpenAI press event, consumer product lead Adam Fry described ChatGPT search volume as “roughly one-fifth (20%) of the volume of queries of Google.” OpenAI didn’t disclose which Google volume estimate it used, making the headline ratio difficult to assess precisely. That missing detail matters because the result depends on what OpenAI included in its count and which measure of Google activity it chose. Without a shared definition and a clear source for the comparison, readers can understand the scale OpenAI is claiming, but can’t independently check the arithmetic.

What counts as a search?

The two services may not be measuring equivalent activity. OpenAI counts a prompt as a search when it triggers a web search. As Fry put it, “Every time you ask a question and go search the web, that’s a search.” That definition makes the 20% comparison dependent on how Google’s queries are counted. A person can ask ChatGPT a question that leads to a web search, while a Google query is generally entered into a dedicated search interface; the numbers may capture related intentions without representing identical user actions. The distinction isn’t a technical footnote. It shapes what the ratio can support.

A query count also says little about the quality or purpose of the activity. Some people may use ChatGPT to explore a subject through follow-up questions, while others may turn to a search engine for a quick fact or a specific website. Similar systems typically serve a mix of tasks, and a single volume figure can’t show how those tasks divide. If you’re working in this space, the useful question is whether people are changing their habits, not just whether a prompt triggered a search.

Pressure on Google, with a revenue question

Even with that caveat, the reported scale points to a new kind of pressure on Google. OpenAI began pushing ChatGPT toward AI-powered search about two years ago; Google has since expanded AI features across Search. The contest now reaches beyond the familiar list of links: both companies are working to answer questions directly, while still connecting users to information on the web. That shift could change what people expect from search, especially when a conversational answer feels more useful than sorting through results.

Google has acknowledged “lower monetization” amid that push, and ChatGPT faces its own challenge in turning search use into revenue. Search volume is only one part of the business question. A service also needs to keep users coming back and find a way to earn from their activity without undermining the experience that drew them in. OpenAI’s reported ratio establishes that ChatGPT search is being used at meaningful scale by its own measure; it doesn’t establish that the activity is profitable or that it will grow at the same pace.

Personalization and the ad trade-off

Fry said most users “do not seem to mind ads” in ChatGPT, adding, “we could be doing this more, actually.” He also said users “really love” the product’s personalization and understanding. Those qualities may help distinguish ChatGPT from a conventional results page, but the reported volume alone doesn’t show how often people return or how the service will monetize that use. Ads can support a free product, yet their placement and frequency can affect whether recommendations feel helpful or commercially driven. And this is the part most people overlook: a user’s tolerance for ads today doesn’t guarantee that tolerance will hold as the product changes.

Here’s the thing: personalization is both a product advantage and a test of trust. A service that remembers context can make a conversation feel more relevant, but it also raises expectations about how carefully the system handles that context. Fry’s comments describe what OpenAI says users value; they don’t provide evidence about retention or about how advertising will fit into those interactions. Promising, but unproven.

What the ratio means next

The 20% figure is best read as a signal of reach, not a verdict on who is winning search. It gives OpenAI a striking way to describe ChatGPT’s scale, while the undisclosed Google benchmark leaves the comparison hard to audit. Future comparisons will be more useful if they make clear what counts as a query and whether they measure the same kinds of activity. Until then, readers should treat the ratio as an indication of momentum, with the measurement caveat attached.

For Google, the number reinforces the case for making Search more useful in an era when people can ask a chatbot to find and explain information. For OpenAI, it raises a tougher follow-up: can the company turn search activity into a lasting habit and a sustainable business without weakening the personalized experience Fry says users value? The numbers here are striking, but they leave those questions open.

More on Google Search:

• Google will disable Search for students unless schools accept AI features in ‘Guided Learning’ push
• Google’s Ocarina of Time easter egg is a blast from the past
• Google AI Mode and AI Overview updates work to elevate links

Follow Ben: Twitter/X, Threads, Bluesky, and Instagram