Zero-click AI answers: what AI summaries do to your traffic

For years, a search engine was a gateway: you asked a question, got ten links and clicked a few. AI summaries change that. The answer sits right on the page, and the links sit below or beside it. In this article, we look at what independent measurements show about the effect on clicks, and at the question that follows: if fewer people click through, what's it worth to be named in the answer?

What Pew Research Center measured

The most careful study comes from Pew Research Center. In March 2025, Pew tracked the browsing of 900 US adults who agreed to have their browsing recorded. About 58% ran at least one Google search that month that produced an AI summary¹.

The result: users who saw an AI summary clicked a regular search result in 8% of visits. Without a summary, it was 15%, nearly twice as often¹. Clicks on the sources inside the summary happened in just 1% of visits. And users were more likely to end their browsing session after a page with an AI summary: 26%, compared with 16% for pages with only traditional results¹.

Pew also shows when these summaries appear. Across the study, 18% of Google searches produced an AI summary. For searches starting with a question word, it was 60%, and for searches of ten words or more, 53%¹. Those are exactly the kinds of queries someone types when they're looking for a provider and describe their problem in a sentence.

One caveat: Pew measures correlation, not cause. Searches with a summary may get fewer clicks for other reasons too. But the pattern is clear, and other measurements back it up.

What happens to the #1 result

Ahrefs compared 300,000 keywords: half with an AI Overview, half informational keywords without one. In March 2025, the presence of an AI Overview correlated with a 34.5% lower click-through rate for the top-ranking page³. A follow-up covering December 2025 reportedly found a 58% drop for the first position⁴. We've seen that update through a secondary source.

Seer Interactive looks at real Search Console data from 42 client organizations and 3,119 informational search terms. For queries with an AI Overview, organic click-through rate fell from 1.76% to 0.61%⁵. Queries without an AI Overview declined too, though less sharply⁵. Bain reaches a similar picture by a different route: about 60% of searches end without a click to another website, and organic traffic drops by an estimated 15% to 25%².

Being named is the new ranking first

Seer's most interesting finding isn't about the decline. It's about the difference inside the AI Overview. Businesses cited in the AI Overview earned 35% more organic clicks and 91% more paid clicks than businesses that weren't cited⁵. Seer is upfront about the limits: strong businesses with already-higher click-through rates may simply be more likely to get cited. Cause and effect can't be separated.

Still, there's a practical lesson here. In a results page with ten links, attention was spread across positions. In an AI answer, it's spread across names: which businesses, sources and websites get mentioned. If you're not in the answer, you have little chance of a click, no matter how well your page ranks. Ahrefs itself notes that AI Overviews often cite several sources, so any single link gets fewer clicks³. According to Pew, 88% of summaries cite three or more sources¹.

Less traffic, better traffic

There's another side to this. The traffic that does come from AI assistants is growing fast and behaves differently. Adobe tracks visits to US online stores. In the first quarter of 2026, traffic from AI sources was 393% higher than a year earlier⁶. In March 2026, that traffic converted 42% better than other traffic, while in March 2025 it still converted 38% worse⁶. By July 2026, the gap had grown to a 60% higher conversion rate, with 53% more revenue per visit⁷.

On-site behavior differs as well. In July 2026, visitors from AI referrals spent 59% more time on site, bounced 33% less and added items to their cart 28% more often than other shoppers⁷. Compared with October 2024, traffic from AI platforms to US online stores had grown 1,219%⁷. That growth rate mostly reflects a low starting point; for most websites, AI traffic is still a small share.

Our reading: people who arrive through an AI assistant have done most of their research already. They've compared options, gotten advice and click through with intent. The volume is smaller than Google traffic, but the visitor is further along in the decision. Note that Adobe's figures cover US online stores, not local service businesses.

Why the click says less than it used to

For most businesses, website traffic has long been the main measure of online visibility. That works less well when much of the decision happens inside an answer nobody clicks. Three things make it tricky:

• A mention without a click is invisible. When an AI model names your business and the customer then looks you up or calls you directly, that shows up as direct traffic, a branded search or a phone call. In your analytics, it doesn't look like AI.
• Referrals aren't always passed along. Some traffic from AI assistant apps arrives without referrer information. Actual AI traffic is likely larger than what analytics tools show. That's our estimate, not something the studies above measure.
• A negative mention costs clicks too. If an AI model names you with a caveat, say about long wait times, a customer may drop off without ever seeing your website. Sentiment in the answer counts.

That's why it makes sense to measure what AI models say about you directly, alongside website traffic: whether you're named, how often, in what context and based on which sources.

What Google and analysts say

Google stresses that AI Overviews drive more than 10% more queries for the types of searches that show them⁸. Gartner, on the other hand, predicted in 2024 that traditional search engine volume would drop 25% by 2026⁹. These views clash less than they seem: people aren't searching less, but a growing share of answers is delivered without a website visit.

What this means for your business

• Website traffic is an incomplete measure. If you get fewer visitors from Google, it may be because the customer already got the answer. The question is whether you were in it.
• Track mentions, not just rankings. In AI answers, what counts is whether you're named, how often, next to whom and in what tone. A Google ranking tells you little about that.
• Sources are the lever. AI answers cite several sources. Knowing which pages and websites get cited tells you where you need to show up.
• Small numbers can be worth a lot. In Adobe's data, AI traffic converts better than average. A few extra mentions in the right answers can be worth more than the visitor count suggests.
• Long questions are the battleground. AI summaries show up mostly for questions and long queries. That's exactly where customers describe their situation ("which installer near me can fit a hybrid heat pump with a rebate?"). Make sure your answer to questions like that is on the web, in plain, factual language.

So the honest picture isn't that search engines are disappearing. It's that value is shifting from the click to the mention. Businesses that only track rankings and visitor counts will notice that shift only when the phone gets quieter.

How GeoReady measures this

GeoReady doesn't track your Google ranking. It tracks whether ChatGPT, Perplexity, Gemini and Claude name you when customers ask their questions. Tracking runs through the official APIs, 3× per week, with repeated runs per prompt. You see your win rate and share of voice against competitors, sentiment with the source sentence, and, per URL, which citations the AI models use, including your own pages. Based on that, GeoReady suggests actions and shows the measured effect after 30 days. So you know whether you're in the answer, even when it doesn't show up in your website analytics.