How Do You Rank in Google AI Overviews? What the 2026 Data Says

If you run a dental group or a home building company and you’ve noticed Google answering your buyers’ questions before they ever get to your site, you’re probably wondering how to get your name into that answer instead of the guy across town. I’ve done this work on my own brand, and the honest answer is a lot less magic than what people are selling you.

What is a Google AI Overview, and how often does it show up?

A Google AI Overview is the AI-written summary that sits above the regular results, with links to the pages it pulled from. A Washington University in St. Louis study published in September 2026 found AI Overviews on 13.7% of all queries but on nearly 65% of question-form queries, with about eight sources cited per overview [1].

So if your buyers type questions, and they do, things like “how much does a custom home cost per square foot in Austin” or “do I need a lawyer after a car wreck,” there’s a good chance an AI Overview is the first thing they read. That’s about eight slots, and your firm is either one of them or it isn’t.

The same WashU team checked 98,020 claims inside AI Overviews and found about 89% were clearly or broadly supported by the sources Google cited [1]. I take that as a reminder that Google mostly says what your page says, so the page has to say the right thing in plain words.

Does Google require anything special to appear in AI Overviews?

No, Google doesn’t require anything special to appear in AI Overviews. Its own documentation says there are “no additional requirements to appear in AI Overviews or AI Mode,” and that a page only has to be indexed and eligible to show in Search with a snippet [2]. The normal SEO basics are what it points you back to.

Those basics are boring but they’re real. Let Google crawl the page, link to it from your other pages so it’s easy to find, keep the important stuff in actual text and not baked into images, and keep any structured data matching what’s visible on the page [2]. Google also says you don’t need new AI text files or special markup [2].

That’s a little deflating if you were hoping for a trick. But it’s good news too, cause it means the work is stuff you control, your own pages, and not some secret setting an agency claims to know about.

Do you need to rank on page one to get cited in an AI Overview?

You don’t need to rank on page one to get cited, but it sure helps. TripleDart found pages in position 1 got cited 55% of the time and pages in position 9 got cited 34% of the time [3]. So ranking is still a head start, and I’d never tell anybody to stop caring about it.

And the old rule that the top 10 owns everything has gotten a lot looser. In Ahrefs’ study of 863,000 keywords and 4 million AI Overview URLs, reported by Search Engine Journal in March 2026, only 38% of cited pages also ranked in the top 10 for the same query, down from 76% in their July 2025 version [4]. YouTube alone made up 5.6% of all AI Overview citations in that data [4].

My read, and it’s only my read, is that Google is answering the follow-up questions too and pulling from whoever answers those well. That’s why a page that covers the real sub-questions a buyer has, cost, timeline, what can go wrong, tends to show up for more than its one keyword. It’s also why I build pages in topic clusters with a pillar page on top, so each of those follow-up questions has a page that answers it.

Where on the page does Google pull its AI Overview answer from?

Google pulls most AI Overview answers from near the top of the page. TripleDart found 35% of citations came out of the first 15% of the page, and the first 100 words, about 4% of a typical cited page, produced 15% of the citations [3]. The passage Google matched ran about 28 words at the median [3].

That 28 words is the part I’d tape to the monitor. It means Google is lifting a couple of sentences, so those couple of sentences have to make sense alone, with your company name and the actual answer in them, not “it” and “this” pointing back at something three paragraphs up.

Don’t chop everything into fragments to get there though. The same study found pages averaging under 12 words per sentence got cited less often [3]. Write like a person explaining it across a desk, just put the answer first.

How to write the top of a page so it can be quoted

1. Put the buyer’s question in the H1, in the words they’d actually type.
2. Answer it in the first two sentences and name your company or the thing you’re explaining, so the answer still works if it’s lifted out alone.
3. Add one specific fact and say where it came from, a price range, a timeline, a number from your own records.
4. Then go deeper, with each H2 written as a real follow-up question and answered in its first sentence.
5. Use a comparison table where the buyer has a real choice to make, since TripleDart found tables on 23.0% of cited top-10 pages against 17.8% of skipped ones [3].

Does schema markup or an llms.txt file help you get into AI Overviews?

Schema markup and AI text files don’t look like an AI Overview lever in the evidence I’ve seen. Ahrefs tracked 1,885 pages that added structured data between August 2025 and March 2026 and found their AI Overview citations fell 4.6% against matched pages that didn’t, a small drop but statistically significant [5].

The same study found AI Mode up 2.4% and ChatGPT up 2.2%, both statistically indistinguishable from zero [5]. And Google says flat out there’s no special schema you need to add and no AI text files you need to create [2].

I still add schema to every page I build, because it helps Google understand who the business is and it can earn rich results. I just don’t sell it as the AI Overview fix, and I’d be careful with anybody who does.

Is being cited in an AI Overview worth anything if nobody clicks?

Being cited in an AI Overview is worth something, but you should be honest with yourself about clicks. Pew Research Center tracked 900 U.S. adults in March 2025 and found people clicked a regular search result in 8% of visits when an AI summary appeared, against 15% when one didn’t, and clicked a link inside the summary in just 1% of visits [6].

So clicks shrink for everybody on those searches. The question is which side of the shrinking pie you’re on. Seer Interactive looked at more than 3,100 queries from June 2024 to September 2025 and found brands cited in AI Overviews got a 35% higher organic click-through rate and a 91% higher paid click-through rate than brands that weren’t [7].

Seer says itself it can’t prove the citation causes the higher clicks [7], and I’d say the same. But being named is the better seat, and the buyer who reads your name in the answer and then searches you by name later doesn’t show up as an AI Overview click anywhere.

This post is part of my larger guide to AI visibility and how buyers find companies through AI tools.

What does it cost to wait on AI Overviews?

The common outcome is slow and quiet. Pew’s numbers say fewer people click through on searches with an AI summary [6], and the WashU numbers say question searches are where those summaries show up most [1]. Put those together and the buyer researching “how long does a dental implant take” reads the cited names first and might never scroll to your listing.

The worst case, a buyer who never hears your name at all, is real on those question searches. It’s not the norm when somebody already knows you and searches your name, because they’ll still find you. So the risk sits in the research stage, where buyers who don’t know you yet build their shortlist.

And it compounds, because every month you wait, the pages Google already trusts for those questions keep getting cited. The next step is cheap. Take the 20 questions your best clients asked before they hired you, search each one in Google, and write down who gets cited. That list tells you exactly where you’re missing.

How I know this works

My own ecommerce brand ranked #1 on Google for its “best” product search, and it was the first product Google’s AI named for that same search when I checked it by hand in July 2026. That same month 18 of its 22 target terms were in the top 3. Its Bing AI citations went from 9 a month in April 2026 to 1,275 a month in August 2026, per the AI performance report in Bing Webmaster Tools.

That’s one brand, and it’s mine, not a client, and it also runs paid ads, so I can’t split out every cause. You can see the results from my own brand and my earlier work, with how each was measured, and the Google side was checked by hand while the Bing side is Bing’s own report. Yours will differ.

Which AI Overview tactics are worth your time?

If you’ve got a small site and somebody in-house who can write, you can do most of this yourself with the steps above, and you should try that first. Where it stops being a weekend project is a site with a few hundred pages across several service lines or locations, which is where a full rebuild starts to make sense.