Someone searched “best budget robot lawn mower” last week and never clicked a single blue link. Google’s AI Overview stitched together a summary, named three brands, and the searcher walked away satisfied. No visit. No pageview. No ad impression for whoever wrote the review that fed the answer.
That’s not a fluke. According to a recent DesignRush roundup, AI Overviews now show up in 43% of Google searches, up from just 15% a year earlier. AI Mode visits have more than doubled in the same stretch. If you write content for a living, or run a site that depends on search traffic, that number should worry you a little. It should also change how you think about what “ranking” even means anymore.
Rankings Are Losing Ground to Citations
For twenty years the game was simple. Get your page to position one, three, or maybe five, and the clicks follow. That model is cracking. When an AI system answers the question directly inside the results page, the click doesn’t happen at all. The Register covered this shift bluntly, citing Pew Research data showing click-through rates roughly halved on queries where an AI Overview appears. Half. Not a dip. A cliff.
So what replaces the click as the thing worth optimizing for? Being named. If an AI system is going to answer the question instead of sending someone to your site, the next best outcome is showing up inside that answer by name. Marketers have started calling this a brand mention, and it’s becoming the currency that backlinks used to be.
Here’s the part that’s easy to miss: a brand mention inside an AI Overview isn’t neutral. It reads like an endorsement. Forbes made this point well in a piece on generative engine optimization, noting that when a name appears inside an AI-generated summary, users tend to treat it as vetted, not just referenced. The AI didn’t just mention the brand. It implicitly vouched for it. That’s a different kind of trust signal than a page-three backlink ever was.
Where the Fight for Citations Is Already Fierce
Some niches figured this out faster than others. E-commerce brands are scrambling. Shopify rolled out tools this summer specifically so merchants can track how their products get surfaced across ChatGPT, Gemini, and Copilot, treating AI platforms as a new kind of storefront traffic source rather than an afterthought. That’s a tell. When a platform as big as Shopify builds infrastructure around AI visibility, the shift has already stopped being theoretical.
Comparison and review sites are feeling it hardest, because their entire business model depends on being the source an AI system pulls from. And nowhere is that scramble more visible than in gambling comparison content, where dozens of near-identical “best sites” pages compete for the same handful of AI citations. Pages built around clear headers, named picks, and scannable structure have a real edge here, because an AI answer engine can lift a clean paragraph and attribute it far more easily than it can parse a wall of unstructured text. That’s exactly why northeasttimes.com’s suggested gambling sites are structured the way they are, with distinct sections and named operators rather than buried mentions, specifically so an AI Overview has something clean to quote and cite.
Gambling involves risk. If you’re reading any comparison content in this space, only wager what you can genuinely afford to lose, and know that BeGambleAware exists if it stops feeling like fun.
That structural discipline matters more than most writers assume. An AI summarizer isn’t reading for style. It’s pattern-matching for extractable claims: a named entity, a specific number, a clear verdict. Content that hedges, rambles, or buries its actual recommendation three paragraphs deep gives the model nothing clean to lift. Content that states its pick early and backs it with a specific detail gets quoted.
Why Publishers Are Panicking, and Why That’s Useful Context
The traffic numbers behind all this are getting genuinely grim for some outlets. Editor & Publisher reported that news publishers across the board are seeing search referrals plunge, and the piece’s blunt framing, that some sites should stop expecting a rebound, isn’t hyperbole so much as a warning. Digiday’s coverage of the so-called zero-click era makes a similar point: publishers of every size are adjusting distribution strategy because the old assumption (write good content, rank well, get visited) no longer holds in the same way.
This isn’t just a media-industry problem. It’s every niche site, every affiliate page, every comparison list. The publishers panicking hardest right now are the ones who were previously the biggest beneficiaries of pure ranking position. That’s the irony. The sites that used to win by being first are now the ones losing the most ground, because “first” doesn’t guarantee a click anymore. It guarantees a chance at being the source text, and that’s a different skill entirely.
What Actually Gets a Site Cited by an AI Overview
Three things seem to matter consistently, based on how these summaries get built.
First, specificity. Vague claims don’t get quoted. A page that says “fast withdrawals” gets ignored. A page that says “withdrawals cleared in under six minutes on three separate tests” gets pulled verbatim, because it’s a concrete, attributable fact.
Second, structure. Headers that match the actual question being asked. Short, declarative sentences near the top of a section. AI summarizers reward content that’s already organized the way an answer would be organized.
Third, and this one gets underrated, consistency of the name itself. If your brand gets referred to five different ways across the web (an abbreviation here, a nickname there, a slightly different spelling somewhere else) the model has a harder time building a confident association between the name and the topic. Clean, consistent naming across a site’s own pages and its external mentions makes citation easier.
None of this replaces good writing. It sits on top of it. A page still has to say something true and useful. But it now also has to say that true, useful thing in a way a machine can lift cleanly, which is a genuinely new writing constraint that didn’t exist five years ago.
The Skill Set Is Shifting Under Everyone’s Feet
Writers who came up optimizing for keyword density and backlink counts are now being asked to optimize for something closer to “quotability.” That’s a real shift in craft, not just tooling. It means shorter, sharper claims. It means naming things explicitly instead of describing them vaguely. It means writing the verdict early instead of building up to it.
Sites that adapt fastest tend to share one habit: they stop trying to rank for everything and start trying to be the definitive, citable answer for a narrower set of questions. That’s a smaller ambition on paper. It performs better in an AI-mediated search result.
The 43% figure isn’t going to shrink. If anything, expect it climbing past half of all searches within the next year or two, based on the trajectory DesignRush laid out. The sites that treat that as an opportunity to get named, rather than a threat to mourn clicks over, are the ones that’ll still matter when the dust settles.
FAQ
What are AI Overviews and why do they matter for SEO? AI Overviews are Google’s AI-generated summaries that appear above traditional search results, answering the query directly. They matter because they can satisfy a searcher without a click, shifting the goal from ranking position to being named inside the summary itself.
How is a brand mention different from a backlink? A backlink passes authority between sites for ranking purposes. A brand mention is simply being named inside an AI-generated answer, with no link required. It builds visibility and implied trust even when no click happens.
Can smaller sites still get cited inside AI Overviews? Yes. Citation depends more on clear, specific, well-structured content than on domain size. A smaller site with a precise, quotable claim can get lifted over a larger site with vague or buried information.
Does this mean traditional SEO is dead? No. Technical SEO and content quality still matter for a page to get crawled, indexed, and considered a credible source. What’s changed is the end goal shifting from click-through to citation-worthiness.
How can a site track whether it’s getting mentioned in AI answers? There’s no single official dashboard yet, though some platforms have begun building tracking tools for this. Manually testing your own target queries across ChatGPT, Gemini, and Google’s AI Mode remains the most reliable check for now.
Either way, the underlying lesson holds. Search hasn’t disappeared. It’s just started answering itself, and the sites that get named inside those answers are the ones that’ll still be standing when the next 43% becomes 60.
