Home Featured How AI Overviews Are Changing Search Behaviour in 2026
FeaturedAIData CenterWriting

How AI Overviews Are Changing Search Behaviour in 2026

Share
Share

Google’s AI Overviews are showing up in more than one out of every five searches across the U.S., and here’s the thing—most searches that get one of these overviews just stop there. People don’t click through to actual websites. SparkToro looked at how Americans were searching in early 2026 and found that 68% of those searches ended without anyone clicking anywhere. When an AI summary popped up, the zero-click rate shot up pretty dramatically. So yeah, websites are definitely losing traffic from this, but it’s not quite as straightforward as just a straight downward slide. Some queries are actually starting to bounce back in terms of clicks, and weirdly enough, brands that get mentioned inside these AI answers tend to see better results on the paid side. What’s really happening is less about whether people are searching at all and more about where they’re getting their answers—and that’s the shift that matters.

Key Takeaways

  • AI-generated summaries appeared on more than 20% of Google searches in the January to April 2026 study period, according to SparkToro’s data as reported by TechNewsDay.
  • Zero-click searches rose from 54% to 72% on queries where AI Overviews appeared, in a study published on SSRN in April 2026, per Search Engine Journal’s coverage.
  • Ahrefs reported in February 2026 that top-ranking pages lost 58% of their click-through rate when an AI Overview appeared, up from 34.5% in April 2025.
  • A preregistered field experiment with about 1,100 participants found that organic clicks to third-party sites fell 39.8% when AI Overviews appeared, with no measurable gain in user satisfaction.
  • Brands cited in AI Overviews earned a 15.74% paid click-through rate on the same queries, versus 11.19% when not cited, according to Seer Interactive’s 2026 data.

What the Numbers Measure

The most-quoted figures in this area measure different things, and treating them as one statistic causes errors. Some studies look at organic click-through rate on queries with an AI Overview. Others measure the share of all searches that end without any click. A third group runs randomized experiments that hide AI features from some users and compare outcomes.

Each method has strengths. Click-through studies use real search console data but cannot prove causation, because AI Overviews tend to appear on certain query types. Zero-click studies capture the whole search session but depend on how the panel is built. Randomized experiments isolate cause but use smaller samples and specific user groups.

Readers should therefore expect a range rather than a single number. The directional finding is consistent across methods: when an AI answer appears, fewer people click through to the sites that supplied the underlying information.

Why Clicks Are Falling

The mechanism is straightforward. An AI Overview places a synthesized answer above the traditional blue links. For simple queries, such as a definition or a current price, users often get what they need without leaving the page. For more complex queries, the summary may satisfy the first question but prompt follow-up searches.

The experimental evidence suggests that the summaries do not necessarily improve the experience. The randomized study found that hiding AI Overviews did not reduce user satisfaction, which raises the question of whether the feature is delivering value to users or simply retaining them on Google’s own surfaces. The researchers wrote that AI Overviews diverted traffic from publishers without delivering measurable improvements in user experience.

Google has not published peer-reviewed or independently verifiable data on click-through rates or publisher referral volumes under AI features. That gap is central to the debate, because publishers and researchers rely on third-party measurement.

The Recovery Signal

Not every study shows a uniform decline. One analysis reported that click-through rates on some searches featuring AI Overviews were beginning to recover, suggesting that users who do click may be more engaged, and that the summaries may send clicks to sources that the AI cites.

This pattern is consistent with a shift in who clicks. Users who click through an AI Overview often have already read a summary and decided they want more detail. That can mean fewer clicks overall but more qualified visits. Seer Interactive’s finding that cited brands earn stronger paid click-through rates points in the same direction: citation inside an AI answer has value even when the organic click is reduced.

Readers should be careful about interpreting recovery. A small rebound on a subset of queries does not reverse a large decline across the search market.

Branded Versus Informational Queries

The effect varies by query type. Informational queries, such as how-to questions and definitions, see the largest losses because they are the easiest for an AI summary to answer. Branded and navigational queries, where the user wants a specific site, behave differently. One agency study reported an 18% increase in click-through rate for branded queries that showed an AI Overview, suggesting that brand recognition can offset the summary effect.

Transactional queries, where the user is ready to buy, are more mixed. Some evidence suggests that AI summaries can reduce paid clicks as well as organic ones on informational searches, but paid performance has held up better on commercial queries. Search marketers should segment reporting by query type rather than looking at a blended average.

Who Is Most Affected

News publishers, reference sites, and content businesses that rely on informational queries face the greatest exposure. Industry data cited by Search Engine Journal indicates that search typically accounts for 20% to 40% of referral traffic to major publishers, and that some sites have seen click-through declines of up to 89% on specific content types.

Smaller sites and niche publishers face a different problem. A site can rank well and still lose traffic if its content is summarized accurately in the AI Overview. Sites with strong brand recognition, original data, or proprietary tools are better protected because users have a reason to visit directly.

What Changes for Content Strategy

Content teams should move from a traffic-first model toward a visibility-and-citation model. The practical steps are not exotic, but they require discipline.

  1. Lead with the answer, and make it extractable. A clear answer in the first paragraph can be cited by AI systems and still earn a visit from users who want depth.
  2. Provide original data. AI systems tend to cite sources that contain information they cannot easily synthesize from elsewhere, such as proprietary surveys, original analysis, and clearly labeled figures.
  3. Build branded demand. Users who already know a source are more likely to click through, even when an AI summary appears.
  4. Track citations, not just clicks. Measure whether your content is cited in AI answers for priority queries, and report that alongside organic click-through rate.
  5. Separate reporting by query intent. Informational, branded, and commercial queries behave differently, and blended metrics hide the shifts.

Measurement Pitfalls

Teams often blame AI Overviews for declines that have other causes. Ranking changes, seasonality, site migrations, and content quality can all reduce traffic. A careful analysis should compare pages that do and do not trigger AI Overviews within the same topic cluster, using consistent time windows.

Teams should also avoid overreacting to single studies. The numbers in circulation vary widely because methodology varies. A sensible approach is to build an internal baseline from your own Search Console data, track it monthly, and adjust strategy when your own data shows a sustained change.

The Policy Question

Publishers have raised concerns about the economics of AI summaries, including whether the search engine should compensate sources whose content powers the answers. Regulators in several jurisdictions have looked at search competition and publisher relations. The policy outcome is not settled, and this article does not forecast it.

For now, the practical reality is that search is shifting from a referral model to an answer model, and content owners must plan for lower baseline traffic from informational queries while protecting the audiences and channels they control directly.

Frequently Asked Questions

How many searches show AI Overviews? SparkToro’s analysis reported that AI-generated summaries appeared on more than 20% of U.S. Google searches during January to April 2026. The share can vary by query type and over time.

Do AI Overviews always reduce clicks? The evidence points to reduced organic clicks on queries where an AI Overview appears, but the size of the effect varies by study and query type. Some queries show recovery or stable performance, particularly branded searches.

Why do zero-click searches matter? A zero-click search ends without any click to an external site. A higher zero-click share means fewer visits for publishers and more of the user journey happening on the search results page.

Are AI Overviews worse for paid search? Some evidence suggests paid clicks on informational queries can also decline, but commercial queries have been more resilient. Results depend on query type and the specific study.

Should I stop investing in SEO? No. Organic search still drives meaningful traffic, and citation inside AI answers depends on the same underlying quality signals. The mix of goals changes, and measurement needs to change with it.

How can I tell if AI Overviews are hurting my site? Compare click-through rates on queries that do and do not trigger AI Overviews, segment by query type, and watch for changes in branded versus non-branded traffic. Treat any single study as context, not proof.

Conclusion: Plan for Lower Clicks, Higher Stakes

AI Overviews are shifting how people search in ways you can actually track. We’re seeing fewer clicks on a lot of informational stuff, and zero-click searches keep going up. The thing is, it’s not hitting everyone equally—the fallout’s pretty uneven. That said, when AI answers cite your work, it does matter. So publishers and marketers really need to dig into their own numbers, break things down by what people are actually looking for, and develop audiences and citations that aren’t totally reliant on that standard blue link anymore.


Discover more from Whiril Media Inc

Subscribe to get the latest posts sent to your email.

Share

Leave a comment

Leave a Reply

Discover more from Whiril Media Inc

Subscribe now to keep reading and get access to the full archive.

Continue reading