Search may be stubborn. Despite the breathless hype surrounding “zero-click” AI search, the predicted exodus of consumers from traditional search bars to chatbots has yet to materialize.
New data suggests that standalone AI tools still command a negligible fraction of the time users spend hunting for information online.
“AI is revolutionizing search, but maybe not in the way that a lot of people think it is,” said Nate Elliott, principal analyst, AI in marketing and commerce at Emarketer, in this video interview with Beet.TV.
“A lot of folks seem to think that consumers are going to stop going to Google and other traditional search engines and spend all their time searching for things on AI tools like ChatGPT and Gemini and Perplexity and Claude. What we’ve found is that that is not yet the case.”
The dominance of legacy search
Emarketer’s analysis, which combined time spent across major generative AI tools against traditional engines, revealed a disconnect in user engagement.
“When you add all that time together, those generative AI tools accounted for only 3.3% of the time people spend looking for information online in the US,” Elliott said. “Traditional search engines like Google and Bing actually account for almost 97% of the time that people spend looking for information.”
This data contradicts forecasts from analyst firms like Gartner, which had predicted a 25% drop in traditional search engine volume by 2026 due to the rise of virtual agents. While the technology is advancing – with worldwide AI spending expected to reach $2.52 trillion in 2026 – consumer habits regarding information retrieval have proven more resistant to change than anticipated.
Blurring the lines of discovery
Elliott and his colleagues have been writing a lot about this space in recent months. Just last week, analyst Grace Harmon wrote: “The promise of AI efficiency may come at the cost of transparency, multi-source information, and the incidental discovery of new publishers and products that link-based search encourages.”
Google has responded to the threat by embedding generative capabilities directly into its core product. Following the launch of Gemini 3 and the more recent Gemini 3.1 Pro, the search giant has increasingly defaulted users into “AI Mode” experiences, effectively keeping the traffic within its own ecosystem rather than losing it to competitors.
“I think they [Google] would like to make AI search just a normal part of regular traditional search,” Elliott said. “They’re working really hard to blur the lines as aggressively as possible. We saw that when they introduced AI overviews a couple of years ago… and we saw it again just a few days ago when they said that people who click to expand an AI overview are basically now going to be taken into an AI Mode experience.”
That integration creates a transition where a user begins with a traditional query but can quickly end up in a conversational interface. “If you want to see that full response from AI overviews, you actually now are in a full-fledged AI chatbot,” Elliott said. “I think sometime soon, they’ll just be defaulting people into AI mode for some searches in some circumstances.”
Optimizing for the new ecosystem
So, for marketers attempting to allocate resources, the ongoing scale of traditional search may the deciding factor.
“If you only have the time, the budget, the resources to focus on search or AI… traditional search is still clearly in the lead,” Elliott said. “30 plus times more time spent with traditional search engines right now than with these AI tools looking for information.”
However, the definition of optimization is shifting as AI overviews take up more real estate. “One of the things we know is that as many as half of Google searches now feature an AI overview,” Elliott said. “If I was trying to optimize for one AI engine over another, I’d be trying to optimize for Google’s tools because five billion people a month do a search on Google.”
