I Checked My Search Traffic and Realized Ranking First Doesn't Mean What It Used To

A few weeks back I was comparing notes with someone who runs a recipe blog, and she said something that made me stop and think. Her site ranks in the top three for a bunch of her target keywords, better than it ever has. Her traffic, meanwhile, has been flat for over a year. That combination used to be impossible. Ranking well and getting fewer visitors at the same time didn't used to make sense. Now it does, and the reason is worth understanding if any part of your income depends on people actually clicking through to your site.

The number that explains everything

Here's the stat that stuck with me most. Back in 2019, roughly half of all Google searches ended without anyone clicking a result at all, the person got their answer straight from the search page and moved on. By this year, that number has climbed to somewhere around sixty five percent, and for searches that trigger an AI Overview specifically, it jumps even higher, up near eighty percent. Ranking in position one on a page that shows an AI Overview reportedly gets clicked on nearly sixty percent less often than it would have a couple years ago.

Put plainly: you can do everything right, write the best answer to a question, earn the top spot, and still watch fewer people actually visit your site, because Google is increasingly answering the question itself before anyone scrolls down to where your link sits.

It's not the same everywhere

What's interesting is how unevenly this hits different kinds of content. If you write about health, general knowledge, or anything informational, "what causes X" or "how does Y work," you're getting hit hard. AI Overviews show up on the vast majority of those searches now, because Google can summarize the answer directly and most people don't need to click anywhere to get it.

Commercial searches, the kind where someone's actually trying to buy something, compare prices, or find a specific business, are barely affected in comparison. If someone's searching for a product to purchase or a local service to hire, they still tend to click through, because an AI summary can't complete a purchase or book an appointment for them. So the honest read here isn't "SEO is dying." It's that the value of ranking for a purely informational question just dropped, while ranking for anything transactional held up much better.

What actually changes if you write content for a living

The instinct a lot of people have is to fight this, to try to trick the system into not summarizing your content, which mostly doesn't work and wastes energy that could go somewhere more useful. The more realistic response is accepting that being cited inside an AI Overview is now a legitimate goal on its own, even without the click, because being named as the source in front of someone actively searching still carries brand value, even if it doesn't carry a visit.

There's a newer discipline people are starting to call answer engine optimization, sitting alongside traditional SEO rather than replacing it. The practical version of that isn't complicated. Answer the actual question clearly and early in the piece, in language a system can lift cleanly, rather than burying the answer three paragraphs down after a long introduction, the exact habit most blogs, including plenty of the advice out there about blogging itself, has trained people into for a decade. Structure matters more now too, clear headings that map to the actual sub-questions someone might ask, because that structure is what makes a page easy to summarize accurately.

And the trust signals that used to be a nice-to-have are becoming closer to a requirement. Where content is coming from, and whether an AI system trusts it enough to cite it in the first place, seems to lean heavily on the same signals search quality raters have talked about for years: real expertise, a genuine author, evidence the person actually knows the subject rather than having assembled it from other pages. Thin, generic content is getting squeezed from both directions now. It doesn't rank well, and even when it does, it's the least likely to get pulled into an AI answer as a trusted source.

What I'd actually do differently

If I were rebuilding a content strategy from scratch today, knowing what this data shows, I'd stop treating "get to position one" as the finish line. Position one on an AI Overview page might send you fewer visitors than position four used to. I'd focus instead on the kinds of questions where someone's clearly about to do something, buy, book, decide, because that's where clicks are holding up. And I'd write the actual answer to the question in the first two or three sentences of any piece, not as some robotic keyword-stuffed opener, just honestly, the way you'd explain it to someone standing in front of you, because that's increasingly the paragraph doing all the work whether a human reads the rest or an AI system quotes it directly.

The recipe blogger I mentioned earlier didn't do anything wrong. The ground just moved under a strategy that used to work perfectly well. That's not a reason to panic. It's a reason to notice that the actual goal, being useful enough that someone trusts what you wrote, hasn't changed at all. Only the path from "useful" to "traffic" has gotten more complicated, and pretending it hasn't is the only real mistake left to make.