If you’ve been using Google with any regularity in 2026, you’ve noticed it. The results page looks different. There’s an AI overview at the top. A lot of queries now return summarized answers that make clicking through to actual websites unnecessary. The experience of searching the web has become, in some ways, more like using Wikipedia — you get an answer, you may not go deeper.
What’s harder to see from the user side is what this is doing to the underlying content ecosystem. And what it’s doing is more consequential than the interface changes suggest.
The Clickthrough Rate Problem Nobody in Publishing Talks About Directly
The data that gets shared publicly about search traffic is carefully curated. Publishers whose traffic has dropped 40% don’t announce it in press releases. But the conversation in private has gotten more candid, and the number that keeps coming up in working groups and publisher forums is the decline in organic clickthrough rates for queries that trigger AI overviews.
Google’s own documentation has acknowledged that AI overviews reduce clicks for some query types. The magnitude varies significantly by topic — informational queries with clear factual answers see the biggest drops; queries where the answer is incomplete, contested, or requires current judgment see much smaller drops. But the direction is consistent.
The Discoverability Concentration Problem
There’s a second effect that’s less discussed: AI overviews tend to cite the same sources repeatedly. A small number of high-authority publications get featured disproportionately often in AI-generated answers, while the long tail of specialized and niche publishers gets cited rarely or never. The effect isn’t malicious — it’s a consequence of models being trained on patterns of what authoritative content looks like, and authoritative content being unevenly distributed.
The result is a feedback loop: publishers with high topical authority grow in discoverability, while the rest shrink. The discoverable surface area of the internet, from a search referral perspective, is getting smaller and more concentrated.
What This Means for the Publishing Business
Publishers who rely on search traffic as their primary audience acquisition channel have had to make a choice: adapt or shrink. The adaptation paths that have shown results in 2025 and 2026 are different than the ones that worked in the 2010s.
First, direct audience relationships have become more valuable than ever. Publishers with large email lists, active subscriber communities, and social channels they own outright have a structural resilience that pure SEO plays don’t provide.
Second, the type of content that drives search traffic has shifted. AI overviews are least disruptive on topics where the right answer is disputed, requires current information, or benefits from specific perspective — where genuine expertise and editorial judgment produce something that can’t be cleanly summarized. The publishers seeing the best search performance in 2026 are the ones producing that kind of content.
Third, the platforms question is being reframed. Rather than asking “how do we rank on Google?”, the more productive question is “where does our audience actually go to find us?”, and then going there directly.
The search experience has changed permanently. The publishers that are doing well in it aren’t the ones that optimized for it best — they’re the ones that stopped depending on it as their primary channel.