Your Rankings Look Fine. Your Traffic Doesn't.
You check Search Console. Position one, same as last quarter. No penalty, no ranking drop, nothing flagged anywhere in your reporting. Then you check your actual traffic, and it's down again.
That gap isn't a glitch in your analytics. It's Google answering the question itself, right there on the results page, before anyone reaches your link.
Ahrefs' analysis of 300,000 keywords, half with an AI Overview present and half without, found that average desktop CTR for the number one position, on searches where an AI Overview shows up, fell from 7.3% in December 2023 to just 1.6% in December 2025. That's a 58% relative decline, at the position every SEO campaign is built around. Position two lost 50.8% of its clicks over the same stretch. Position three lost 46.4%. Even position ten, sitting at the bottom of page one, still lost 19.4%.
Ranking well used to be the whole game. Now it's necessary, and no longer sufficient.
The Zero-Click Search Is the New Normal
This isn't isolated to a handful of keywords that happen to trigger an AI Overview. SparkToro's analysis of Similarweb clickstream data found that 68.01% of all US Google searches ended without a single click during the first four months of 2026. Two years earlier, in 2024, that figure was 60.45%, a jump of more than 7.5 percentage points in just two years, in a market that was already mostly zero-click to begin with.
Put plainly, less than a third of Google searches in the US now send anyone anywhere. The rest end on the results page, frequently with an AI-generated answer sitting above every organic result.
That's a different problem than not showing up at all. A page that isn't ranking on Google has a visibility problem you can usually diagnose and repair with the fundamentals, technical health, content that matches search intent, authority signals. A page that ranks fine and is still losing traffic has an attention problem. Google is keeping the click for itself.
What Actually Still Moves the Needle
If rank position alone doesn't guarantee a click anymore, what does?
Getting Cited Beats Just Ranking
Seer Interactive tracked 53 brands across 5.47 million queries and 2.43 billion organic impressions between January 2025 and February 2026, and the gap it found is the clearest actionable signal in this whole shift.
Per 1,000,000 organic impressions on informational queries, pages saw roughly 33,500 clicks when no AI Overview appeared at all. When an AI Overview did appear and the brand was cited inside it, clicks landed around 20,743, a real drop from the no-AIO baseline, but still healthy. When an AI Overview appeared and the brand was not cited, clicks fell to just 9,445, roughly 2.2 times fewer than being cited earns. Ranking below an AI Overview that quotes someone else is close to invisible.
Ranking Top 10 Is a Weaker Guarantee Than It Was a Year Ago
It used to be reasonable to assume that ranking on page one meant you'd eventually get cited too. Ahrefs found that 76% of AI Overview citations came from pages ranking in Google's top 10 as of July 2025. A follow-up study just eight months later, in March 2026, found that share had fallen to 37.9%. Google's move to Gemini 3 in January 2026 pushed AI Overviews to lean much harder on expanded "query fan-out" sub-searches that never surface on the visible top-10 results page at all. Ranking well still helps. It just isn't the guarantee of AI visibility it was twelve months ago.
Recency Is a Real Lever
Seer Interactive's analysis of more than 5,000 cited URLs with extractable publish dates found that roughly 44% of Google AI Overview citations in 2025 came from content published that same year, and about 85% came from content published between 2023 and 2025. If your best page on a topic hasn't been touched since 2022, it's competing against content Google considers current, and it's losing on that basis alone, regardless of how good the writing is.
Word Count Is Not
One thing that isn't a lever, no matter how often it gets repeated as advice, is length. Ahrefs analyzed 174,048 pages and found essentially no correlation between content length and AI Overview citation, a Spearman correlation of just 0.04. More than half of cited pages, 53.4%, run under 1,000 words. The average cited page sits at 1,282 words. Padding a page to hit some arbitrary word count target does nothing for citation odds. Answering the question well does.
Schema Markup Won't Fix This Either
A common piece of advice going around right now is to slap FAQ or Article schema on every page and call the AI visibility problem solved. It's worth being direct about it, that advice doesn't hold up.
A March 2026 Ahrefs controlled study tracked 1,885 pages that added JSON-LD schema against a 4,000-page control group and found no meaningful citation lift on ChatGPT, up just 2.2%, statistically nothing, or on Google's AI Mode, up 2.4%, also statistically nothing. On Google AI Overviews specifically, schema pages actually saw a small measured decline of 4.6%. Add schema because it's good technical practice and it helps machines parse your page accurately. Don't add it expecting more AI citations. The data doesn't support that expectation.
What This Means for Your Business
Rankings still matter. They're the floor, not the ceiling anymore. A page that doesn't rank at all was never going to get cited or clicked either way, and that's the separate problem covered in why your website isn't ranking on Google. But a page that ranks well and still isn't earning traffic is dealing with something newer. Google increasingly answers the question itself, and only sometimes credits the source.
The businesses handling this well aren't chasing position one any harder. They're making their content genuinely worth citing, kept current, structured so an answer engine can lift a clean and accurate answer straight from the page, and built on real expertise rather than padded word counts or schema shortcuts. That's the actual shift, from ranking for a click to being the answer that gets credited when Google or an AI Overview needs one.
That's the discipline behind Answer Engine Optimization. Our deeper breakdown of what AEO actually is walks through how citation-focused optimization differs from ranking-focused SEO in more depth. Paired with a solid search engine optimization foundation, that's how a business keeps its rankings and its traffic moving in the same direction again.
