Can you appear in an AI Overview without ranking on page 1?

Yes. Ahrefs' March 2026 study of 4 million AI Overview URLs found that only 37.9% of citations rank in Google's top 10, down from about 76% in July 2025. The other 62% split almost evenly between pages ranking 11 to 100 and pages that do not rank in the top 100 at all. Most citations now arrive through query fan-out, not the visible page-1 SERP.

What is query fan-out?

Query fan-out is the mechanic an AI engine uses to break one user prompt into several sub-queries, retrieve sources for each sub-query separately, and fuse the results into one generated answer. Google has confirmed in its own developer documentation that AI Overviews and AI Mode use query fan-out. A citation can come from any of those sub-query retrievals, not only from the page that ranks for the user's exact words.

How many sub-queries should I map for one target question?

Map 6 to 20 sub-queries per target question. Pull them from People Also Ask boxes, Perplexity's related-questions panel, Reddit and forum threads on the topic, and a fan-out visibility tool such as Profound. Fewer than six usually misses real facets of the question. More than 20 spreads your writing effort too thin to build a real passage for each one.

Do I need FAQ or HowTo schema to win a fan-out citation?

Schema is not required, but it lowers the friction for an engine to parse a passage as a direct answer. FAQPage schema marks a single question-and-answer pair explicitly. HowTo schema marks a numbered sequence of steps. Add whichever matches the shape of the sub-query you are targeting, and treat schema as structural hygiene that supports the passage, not a replacement for writing the direct answer itself.

Which tool should I use to see the sub-queries an AI engine generates for my target question?

SE Visible and Semrush track whether an AI Overview appears and whether you are cited, at a query level and at scale, which makes them the right first check. Profound goes a step further and surfaces the actual fan-out sub-queries an engine generates for a prompt. Temso covers citation tracking across engines plus the content and schema fixes to close a gap, in one plan from $89 a month.

How to Get Cited in an AI Overview for a Query You Don't Rank On (The Fan-Out Route)

Q: What does Search Console's Generative AI Performance report actually show?

It shows impressions: how often a specific page appeared inside Google AI Overviews or AI Mode, broken down by page, country, device, and date. It does not show which query triggered the impression, whether anyone clicked, or where your citation sat inside the answer. Use it to confirm a page is being pulled into AI Overviews at all, then pair it with a third-party citation tracker to confirm the actual sub-query and citation.

Last updated September 2026.

You built the page. You tightened the title tag, the headers, the internal links. You still sit on page two for the exact phrase your buyers type into Google. So you assume the AI Overview is closed to you too.

It isn’t.

An AI Overview doesn’t read a search the way you do. It reads the question behind the question, and it retrieves sources for each piece separately. A citation can land on your domain even when your page never cracks the top 10 for the query you were chasing.

This playbook shows you how to find that opening and build the one passage that wins it.

Why page-1 ranking stopped being the ticket in

For years, the shortcut worked: rank in the top 10, and an AI Overview citation followed most of the time. Ahrefs tested that assumption twice, a year apart, and the second test tells a different story.

In July 2025, Ahrefs analyzed 1.9 million AI Overview citations and found that 76.1% of them came from pages ranking in Google’s top 10. Rank well, get cited. The math was simple.

Then, in March 2026, Ahrefs ran the study again at a much larger scale: 863,000 keyword SERPs and 4 million AI Overview URLs, pulled from its Brand Radar tool. The result flipped. Only 37.9% of citations still came from top-10 pages. The remaining share split almost evenly between pages ranking 11 to 100 (31.2%) and pages that didn’t rank in the top 100 at all (31.0%).

That single change rewrites the playbook. Ranking well still helps. It just isn’t the only way in anymore, and for a growing share of citations, it isn’t the way in at all.

The fan-out mechanic, in one picture

Here’s what happens between the moment someone types a question and the moment an AI Overview shows an answer.

        Your target query:
"can you appear in AI Overviews
    without ranking on page 1"
                |
                v
     The engine fans it out
                |
    --------------------------------------
    |        |        |        |        |
  sub-Q1   sub-Q2   sub-Q3   sub-Q4   sub-Q5
    |        |        |        |        |
    v        v        v        v        v
  pages    pages    pages    pages    pages
retrieved retrieved retrieved retrieved retrieved
  for Q1    for Q2    for Q3    for Q4    for Q5
    |        |        |        |        |
    ---------------- fuse into one --------------
                      |
                      v
            One generated answer,
         citations pulled from whichever
            pages won each sub-query

Notice what’s missing: your original query, answered by one ranked page. That page still exists in Google’s index. It’s just no longer the only door into the answer. Five separate doors open, and you only need to win one of them.

The 8-step playbook

Each step builds on the one before it. Work through them in order the first time, then loop back to Step 2 for the next sub-query once you’ve won your first citation.

Step 1: Confirm the AI Overview actually triggers for your query

Before you map anything, check that an AI Overview shows up at all for your target query, and check who currently holds the citation. A tool built for this beats manual spot checks: SE Visible and Semrush both track AI Overview occurrence and citation status at the query level, at a scale that repeated manual searches can’t match.

Run your target query through one of them first. No AI Overview, no fan-out opportunity yet. An AI Overview with a competitor cited and you absent is exactly the gap this playbook closes.

Step 2: Map 6 to 20 sub-queries behind your target question

This is the step that separates the fan-out route from ordinary keyword research. You aren’t listing synonyms of your target query. You’re listing the distinct sub-questions an engine would ask on its way to answering it.

Pull candidates from four places:

People Also Ask boxes on your target query’s own SERP.
Perplexity’s related-questions panel, shown after any search.
Reddit and forum threads where people discuss the topic in their own words.
A fan-out visibility tool. Profound publishes research specifically on this mechanic, and its platform surfaces the actual sub-queries an engine generates for a prompt, which beats guessing from People Also Ask alone.

Aim for 6 to 20 sub-queries. Fewer than six usually means you stopped too early and missed real facets of the question. More than 20 spreads your writing effort too thin to build a genuine passage for each one.

Use a simple worksheet to keep the map organized:

#	Sub-query the engine likely generates	Facet it covers	Priority	Owned content?
1	Does an AI Overview citation require a top-10 ranking?	The core myth to correct	High	Build
2	How does Google AI Overview choose its sources?	Retrieval mechanics	High	Existing post or new section
3	What is query fan-out in AI Overviews?	Definition	Medium	Link to an existing explainer
4	Why do AI Overviews cite pages that don’t rank on page 1?	The supporting data	High	Build
5	How do I track AI Overview citations without a rank tracker?	Measurement	Medium	Build, plus a tool mention
6	What tools show the sub-queries behind an AI Overview?	Tool discovery	Low	Build

Fill in your own version with your actual target question. The columns matter more than the example rows: sub-query, facet, priority, and whether you already own a page that answers it.

Step 3: Pick one sub-query to own first

Don’t try to cover all 6 to 20 sub-queries in one pass. Pick the highest-priority row from your worksheet and build one passage that answers it better than anything currently retrieved.

Good first picks share three traits: the facet isn’t already covered anywhere on your site, competitors haven’t answered it directly either, and you can answer it more precisely than a generic source would. One strong passage beats five thin ones.

Step 4: Build the passage: H2-as-question, 60 to 80 word answer

Write the sub-query as your heading, word for word or close to it. Answer it in the next 60 to 80 words, with no throat-clearing before the answer starts.

Here’s what that looks like using row 1 from the worksheet above.

Does an AI Overview citation require a top-10 Google ranking?

No. Ahrefs’ March 2026 study of 4 million AI Overview URLs found that only 37.9% of citations came from pages ranking in Google’s top 10, down from 76.1% in July 2025. The rest were pulled from pages ranking 11 to 100, or from pages that didn’t rank in the top 100 at all, through the fan-out sub-queries behind the question.

That block is self-contained. An engine can lift it whole and it still makes sense, because it doesn’t lean on a sentence three paragraphs earlier for context.

Step 5: Mark up the passage with FAQPage or HowTo schema

Schema doesn’t guarantee a citation, but it removes ambiguity for the engine parsing your page. Match the type to the shape of your passage.

A single question-and-answer pair like the one above fits FAQPage:

{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [
    {
      "@type": "Question",
      "name": "Does an AI Overview citation require a top-10 Google ranking?",
      "acceptedAnswer": {
        "@type": "Answer",
        "text": "No. Ahrefs' March 2026 study of 4 million AI Overview URLs found that only 37.9% of citations came from pages ranking in Google's top 10, down from 76.1% in July 2025. The rest were pulled from pages ranking 11 to 100, or from pages that didn't rank in the top 100 at all, through the fan-out sub-queries behind the question."
      }
    }
  ]
}

A sub-query that’s really a procedure, “how do I check whether I’m cited,” fits HowTo instead. This whole playbook is a working example:

{
  "@context": "https://schema.org",
  "@type": "HowTo",
  "name": "How to get cited in an AI Overview for a query where you don't rank on page 1",
  "step": [
    { "@type": "HowToStep", "position": 1, "name": "Confirm the AI Overview triggers for your query" },
    { "@type": "HowToStep", "position": 2, "name": "Map 6 to 20 sub-queries behind the question" },
    { "@type": "HowToStep", "position": 3, "name": "Pick one sub-query to own first" },
    { "@type": "HowToStep", "position": 4, "name": "Write an H2-as-question with a 60 to 80 word answer" },
    { "@type": "HowToStep", "position": 5, "name": "Mark up the passage with FAQPage or HowTo schema" },
    { "@type": "HowToStep", "position": 6, "name": "Publish the passage where it can be retrieved standalone" },
    { "@type": "HowToStep", "position": 7, "name": "Check Search Console's Generative AI Performance report" },
    { "@type": "HowToStep", "position": 8, "name": "Confirm the citation and repeat for the next sub-query" }
  ]
}

Step 6: Publish the passage where it can be retrieved standalone

Add the passage to a page that will actually get crawled and reindexed soon: an existing FAQ section on your target topic, or a new page if nothing close to it exists yet. Either works, as long as the passage doesn’t depend on context from earlier on the page to make sense. A retrieved passage has to stand on its own, because the engine may only pull those two or three sentences, not the paragraph before them.

Step 7: Verify with Search Console’s Generative AI Performance report

Once the page is live and reindexed, open Search Console’s Generative AI Performance report and filter to the specific URL you just published. Watch for impressions to start registering there over the following days to weeks.

Read the result carefully. The report groups impressions by page, country, device, and date. It has no queries dimension, so a rise in impressions tells you the engine started pulling that page into AI Overviews or AI Mode. It does not tell you which sub-query triggered it, and it does not confirm you were named as a citation rather than used as silent grounding.

That gap is why this step is a checkpoint, not a finish line. For the full method behind reading this report, see Search Console’s Generative AI Performance report, explained.

Step 8: Confirm the citation and repeat for the next sub-query

Run your specific sub-query through a citation tracker to confirm the win directly. SE Visible, Semrush, or Profound will all show whether your domain is now cited for that exact sub-query, not just whether the page picked up impressions.

If you’d rather run the citation check and the content and schema fixes from one workflow instead of stitching tools together, Temso covers that loop end to end. It tracks citations across ChatGPT, Perplexity, Google AI Overviews, Gemini, and Microsoft Copilot, and it includes an AI SEO agent that can draft the next passage for you, on a flat plan from $89 a month with unlimited projects, users, and recommendations.

Once you’ve confirmed the citation, go back to your worksheet from Step 2 and start the next row. Winning a fan-out citation isn’t a one-time project. It’s a queue you work through, one sub-query at a time, until you’ve covered the facets that matter most.

Where each tool fits in this playbook

Tool	Fits best at	Why
SE Visible or Semrush	Step 1 and Step 8: confirm and verify	Track AI Overview occurrence and citation status per query, at scale
Profound	Step 2: map the sub-queries	Surfaces the actual fan-out sub-queries an engine generates, not just guesses from People Also Ask
Temso	Step 8: verify and close the gap	Citation tracking plus content and schema fixes in one platform, from $89/mo

None of these tools writes the passage for you unless you ask them to. They tell you where the gap is and whether you closed it. The writing in Steps 4 and 5 is still yours to do well.

Start with your own worksheet

Pick one query where you don’t rank on page 1 but wish you were cited. Run Step 1 today, build your sub-query map this week, and publish your first passage before the month is out. The full tool comparison, including SE Visible, Semrush, Profound, and Temso side by side on pricing and coverage, is at /rankings/aeo-tools. Core terms like citation, query fan-out, and share of voice are defined in the glossary, and the scoring criteria behind every tool mention on this site are at /methodology.