askadia

Guides · Updated November 19, 2026

How to show up in Google AI Overviews: get your page cited (260 overviews on our own product)

The short answer: rank on Google for the search, with a page written for that exact question, so the AI Overview cites it. Being cited is how you get named: across 260 AI Overviews on our own product's buying questions, when the overview cited one of our pages it named the product in 139 of 153 answers (91%); when it didn't, in 12 of 107 (11%). The pages it cited were blog posts that answer the question, far more than the homepage. Where we had no such page, we weren't there.

What we measured

We make PodSeeker, a podcast database for PR teams, and track it with Askadia. Its 20 non-branded buying questions (“best tool for a podcast booker to find contacts”, “alternatives to Podchaser”, “how do I find podcasts that are booking guests”) were searched on Google, logged out, from the US, 13 times each between October 31 and November 18, 2026, and we read every AI Overview: 260 overviews, with the pages each cited and Google's results for the same search.

Rank first: AI Overviews read what ranks

An AI Overview answers the search as typed, from pages Google already ranks, and runs a few related searches of its own (Google calls it query fan-out). Google says there's nothing extra to do: a page has to be indexed and eligible for a snippet, and the rest is ordinary SEO (Google Search Central). Our numbers agree. Over two weeks to November 14, we put PodSeeker's Google rank for each search beside whether that search's overview named it:

PodSeeker's Google rankThe AI Overview named it
#151 of 56 answers (91%)
#2–1033 of 69 (48%)
Not in the top 1030 of 75 (40%)

591 of the 993 sites those overviews cited were in Google's top 10 for the same search. Large studies pooled across millions of searches find less, and disagree with each other: 37.9% of cited pages in the top 10 (Ahrefs, 863,000 searches, March 2026), 43% (Conductor, 168 million citations, summer 2026), where the #1 result was cited 24.9% of the time and the #10 result 10.2%. On a product's own buying questions, ranking first was most of the work. It's not all of it: below the top 10 PodSeeker was still named in 40%, through the related searches and through other people's pages.

Get your page cited: that's how you get named

Split the 260 overviews by whether they cited one of PodSeeker's pages:

The AI Overview…Named PodSeeker
cited a PodSeeker page139 of 153 (91%)
cited none12 of 107 (11%)

Put the other way: of the 151 overviews that named PodSeeker, 139 cited its own page. AI Overviews rarely named it from what others wrote; they named it when they read it from us. That's correlation (a page that ranks gets both), but it says where to put the work: a page of yours that ranks for the question, and says plainly what your product does for the person asking.

Which of your pages get cited

Not the homepage, mostly. Of the 193 times an overview cited PodSeeker, 161 were blog posts and 19 the homepage. Each question was won by the page written for it:

Question searchedPage cited mostNamed
What's the best tool for a podcast booker to find podcast contacts for clients?“How to contact podcast hosts” (11 of 13)12 of 13
Is Muck Rack enough for podcast guest booking, or does our PR team need a separate podcast tool?“PodSeeker vs Muck Rack” (12 of 13)11 of 13
What are good alternatives to Podchaser for a small PR agency…?“PodSeeker vs Podchaser”, “Podchaser alternatives” (4 each)13 of 13
How should a PR team prioritize podcasts when pitching several clients?“How to decide if a podcast is a good fit” (8 of 13)9 of 13
What's the best podcast guest pitching tool for a solo consultant…?none written for it (cited once)0 of 13

Comparison and alternatives pages (“X vs Y”, “X alternatives”) won the questions that name a competitor, and how-to posts won the how-to questions, with the product named inside the answer. A how-to post isn't always enough: on “How do I get my company's executives booked on podcasts?”, PodSeeker's post was cited in 7 overviews and the product named in 3; an overview answering “how do I” names fewer products.

YouTube, Reddit and other people's lists

The overviews cited 6.2 pages each, and not only websites. YouTube videos were cited in 128 of the 260 and Reddit threads in 88. The top three videos were cited in 26, 23 and 16 overviews: one on how booking agencies pitch, one on ways to find podcasts to guest on, and a tutorial for a competitor, Rephonic. On our public AI visibility tools questions YouTube was the most-cited site too, and Ahrefs finds it the most-cited domain in AI Overviews overall (Ahrefs). So a short video that answers the question, on the channel your buyers search, is a second way in, and a video of yours is your page. Reddit threads are other people's: take part where your product genuinely fits, and say who you are.

Named isn't picked

AI Overviews list; they rarely choose. PodSeeker was the overview's top pick in 15 of 260, even though it came first among the products in 63 of the 151 it was named in, and where it ranked #1 on Google it was the top pick in 3 of 56. ChatGPT is the opposite: on our 40 AI visibility tools questions it named a top pick in 28 answers, AI Overviews in 4. Track being named and your place in the list; don't wait for a recommendation Google rarely makes. More in ChatGPT vs Google AI Mode vs AI Overviews.

What triggers an AI Overview?

Google decides per search, and the same search can get one one day and not the next. In the largest study, AI Overviews showed on 20.5% of US searches, most often on long questions (46.4% at seven words or more) and much less on commercial ones (4.3%) (Ahrefs, 146 million results, September 2025). That's moving: commercial searches with an AI Overview grew 71% between November 2025 and April 2026 (Semrush, 600,000 keywords). Buying questions written the way buyers ask, long and specific, are the kind that get one: every one of our 40 public questions about AI visibility tools did.

How to show up in AI Overviews

  1. List the questions buyers search before they choose, in their words: “best X for Y”, “X alternatives”, “X vs Y”, “how do I do the job X does”. Search each one and see whether it shows an AI Overview, and which pages it cites.
  2. Write one page per question, a comparison, an alternatives page or a how-to, that answers it in the first lines and says what your product does for that person. On our questions, the page written for the question was the page cited.
  3. Rank it. Indexed, crawlable, linked from your other pages, fast; the usual SEO. Where PodSeeker ranked first, the overview named it 91% of the time.
  4. Answer it on YouTube too, where your buyers search: a video was cited in about half the overviews.
  5. Get onto the pages it cites that aren't yours: the lists and threads that keep coming up. Ask the author; never pay for a place.
  6. Check it more than once. Overviews change from day to day; count how many of a question's overviews name you, not one search.

Skip what has no evidence behind it: special schema or AI text files (Google says neither is needed), and rewriting pages for “AI” instead of for the question. More on Google's other AI in How to rank in Google AI Mode, which works differently.

Where am I cited in AI Overviews?

Search Console counts AI Overviews inside ordinary web results, and its generative AI report shows impressions without the searches, so it can't tell you which searches named you or which page an overview cited. To know, search each question logged out, several times, and read the sources, or let a tracker do it. Askadia (ours) does it on a schedule: each question searched on Google, every AI Overview word for word with the pages it cited and Google's top 10 beside it, whether it named you and cited you, counted in questions. The AI Overviews tracker is 2,000 answers a month for $29, ChatGPT and Google AI Mode included.

With your agent, in Askadia (ours)

Give Claude Code, Cursor, Codex or any agent one line, Read https://askadia.ai/docs/api.md, and approve it once, or add Askadia's MCP server (in Claude Code or Cursor, or as a connector in Claude.ai or ChatGPT). Then ask it to:

  • Search your buyers' questions on Google AI Overviews on the days you choose, and tell you on which ones the overview names you, cites you, or neither.
  • Read each overview with Google's top 10 beside it, and find the questions where you rank but aren't cited, or have no page at all.
  • List the pages AI Overviews cite most for your questions (Sources), and mark your own videos and listings elsewhere as yours, so they count as you.
  • Turn the lists and threads that leave you out into todos, and draft the note to each author for you to send.

Included in the $29 plan, and reading is unlimited. It proposes changes until you tell it it can act, and the plan and its answers stay your click.

That's how we got the numbers above: PodSeeker's agent works through Askadia's API, and this study is its answers' export.

Questions

What are AI Overview sources?

The pages Google lists beside an AI Overview, and links inside it: mostly pages ranking for the search or its related searches, plus YouTube videos and Reddit threads. On PodSeeker's questions an overview cited 6.2 on average.

Does being cited mean being recommended?

Not quite, but close: on our questions, an overview citing our page named the product 91% of the time. It was the top pick far less often (15 of 260 overviews): Google lists options more than it picks.

Do AI Overviews still send clicks?

Fewer. In Pew's panel, people clicked a result in 8% of visits with an AI summary against 15% without, and a link inside the summary in 1% (Pew Research Center, 900 adults, March 2025). Being named still puts your product in front of a buyer who may search for it next.

Is an AI Overview the same as Google AI Mode?

No. AI Mode is the chat tab: it runs its own searches, cites about 3 pages and leans on a few lists; your Google rank predicts it much less. On our 40 questions the two never shared a top pick.

Can I stop my pages being used in AI Overviews?

Yes, with nosnippet, data-nosnippet or max-snippet, which also limit your ordinary snippets (Google). Blocking Google-Extended doesn't affect Search.

How we got these numbers

PodSeeker's 25 questions are tracked on ChatGPT, Google AI Overviews and Google AI Mode; the 5 that name PodSeeker are left out here. Each was searched on Google logged out, from the US, 13 times between October 31 and November 18, 2026 (one counted answer per question per day). “Cited” means one of the overview's sources is on podseeker.co; “named” means its text names PodSeeker, read the same way for every answer. The rank table is the same questions over the two weeks to November 14, each overview counted under the rank Google showed for that search that day. Askadia is an AI visibility tracker, and PodSeeker is our own product.

Related: ChatGPT vs Google AI Mode vs AI Overviews · How to rank in Google AI Mode · How to rank in ChatGPT

See whether AI Overviews name your product

The free checkup searches 10 questions your buyers ask and shows each AI Overview, the pages it cited and whether it named you, with ChatGPT and AI Mode beside it, in about 2 minutes.

100% free: no email, no sign-up, no credit card · Takes about 2 minutes