Google's AI Answers Now Eat Most of the Clicks, and Chatbots Send Back Under 1%. If You Run a Content Site Solo, Stop Renting Distribution From Search.

I've written twice in the last month that Google is replacing the link with the answer: once when it pushed AI Overviews to the top of the page, once when it started summarizing articles inside Chrome. Each time the data was bad. Now it's worse, and it's coming from sources I'd cite without flinching.

An Ahrefs study of hundreds of thousands of keywords found that when an AI Overview appears, the top organic result's click-through rate falls by roughly 58%. Small publishers are reporting referral declines in the neighborhood of 60%. And the thing that was supposed to replace search traffic (AI chatbots sending people to your site) still accounts for under 1% of publisher page-view referrals. If your content business runs on Google sending strangers to your pages, the channel isn't softening. It's structurally getting smaller.

The number that ends the old playbook

For fifteen years the content playbook had one move: rank higher. Better than the result below you meant more clicks, and more clicks meant more business. That loop is what SEO was.

The AI Overview breaks the loop at the source. When Google answers the question at the top of the page, the user got what they came for before they ever reached a blue link. Ahrefs measured the top result's organic CTR collapsing: in one widely cited example, a major publisher's desktop CTR fell from around 25% to under 3% when an AI Overview surfaced above its link. You can be the number one result for a query and still lose most of the traffic you used to get from being number one, because the click that used to be automatic now happens only when the answer box didn't satisfy.

"Rank higher" doesn't fix that. You already ranked. The destination moved.

Why the chatbot referrals don't save you

The optimistic story for the last year has been that AI assistants would replace the lost search clicks: people would ask ChatGPT, it would cite you, they'd come read. The citations are real. The traffic isn't, not at any scale that matters. Chatbots still send back under 1% of referrals across publishers. Even large year-over-year growth in chatbot referrals is growth on a tiny base; it's nowhere near enough to offset what AI Overviews are taking off the top.

So planning your business around "I'll get found in the AI answers instead" is planning around a channel that, today, returns almost nobody. Maybe that changes. You can't pay your hosting bill with maybe.

The pivot: own the next reader, don't rent them

Here's the shift I'm actually making, and it's less dramatic than "SEO is dead." SEO isn't dead. It's demoted: from a growth channel you can build a business on to a weather system you benefit from when it's good and can't depend on when it's not.

The thing you build the business on instead is distribution you own. Concretely, for a solo content operator, that means a few unglamorous things. An email list, because it's the one channel where you reach the reader directly without an algorithm deciding whether you're allowed to. A direct audience somewhere you post and people actually see it. And products or offers that don't require a fresh click from a stranger to make money: things your existing readers buy, not things a search visitor stumbles into.

None of that is new advice. What's new is that the alternative ("just rank and the traffic comes") has stopped being a real option for most niches. The owned-distribution work used to be the thing you'd get to once SEO plateaued. Now it's the thing, and search is the bonus on top.

The honest counter-take

I'm describing an average, and you don't run an average site. Some niches are barely touched by AI Overviews yet: queries where Google is cautious about answering directly, where the intent is transactional, where the user genuinely needs to land on a specific page. If that's your category, your search traffic may be holding fine and this reads as alarmism. Check your own analytics before you torch your content strategy; the aggregate decline doesn't tell you what's happening to your pages.

And owned distribution is hard in a way ranking never was. An email list grows slowly and demands you have something to say on a schedule. It's more work than writing for the algorithm. I'm not pretending the pivot is free. I'm saying the bill for not making it is now larger than the bill for making it.

What I'd actually do

Start with your own numbers, not mine. Pull the last twelve months of search traffic by page and find out whether you're in the niche that's getting hit or the one that's holding. That tells you how fast you need to move.

Then, regardless of the answer, start the owned channel now while you still have the search traffic to seed it. Put an email capture on your best-performing pages and give people a real reason to hand over the address: the thing they'd actually want, not a generic "subscribe for updates." Every visitor search still sends you is a chance to convert a rented reader into one you own. The mistake isn't losing search traffic; that's happening to everyone. The mistake is watching it fall and not using what's left to build the channel that doesn't depend on Google's mood.

I run a content site. I'm treating search as weather now, and I'm spending the energy I used to put into chasing rankings into the list and the audience I actually control. That's the only distribution a one-person operation can count on through whatever the answer box does next.

Author

Lukas

@lukcombinator

Sources

• How AI Overviews affect click-through rates (Ahrefs study coverage)
• Google AI Overviews linked to publisher referral traffic drop (Digiday)
• Google Search referrals to the web have plummeted; AI links 'less than 1%' of traffic (9to5Google)
• Small publishers losing referral traffic in the AI era (Android Headlines)