On this page · 11 sections
- What Google has actually said
- How AI Overviews and AI Mode choose pages
- Step-by-step: how to get a page cited in AI Overviews
- What you can ignore, according to Google
- Snippet controls: nosnippet, max-snippet and data-nosnippet
- How to measure AI Overview and AI Mode visibility in Search Console
- What third-party citation studies say
- India notes for AI Overviews and AI Mode
- For IT teams and publishers
- How eCorpIT can help
- References
Summary. To get a page cited in Google AI Overviews and AI Mode, do what ranks a page in Google Search: make sure it is indexed and eligible to show a snippet, and make it the most useful, original source on its sub-topic. Google's documentation says these features have no extra requirements: no new machine-readable files, no AI text files and no special schema.org markup. Nor is there an opt-in: inclusion is the default in Search Console's Search generative AI control, which reached every site worldwide on 31 August 2026 and exists to exclude a site; an exclusion usually takes effect within 1 to 2 days. Both features ground answers in Google's index and use query fan-out, running related sub-queries, so a page that answers one precise sub-question can be cited without ranking for the main query. In Ahrefs' July 2025 study 76% of cited pages ranked in the top 10; its March 2026 update, with a revised method across 863,000 results pages, found 38%. Pew Research Center's March 2025 US data showed users clicked a link inside an AI summary on 1% of visits to pages that showed one. To limit use, nosnippet blocks a page as a direct input, max-snippet caps it and data-nosnippet hides sections. Measure impressions in Search Console's generative AI performance report.
About this article. For site owners, SEO leads, content teams and developers who want pages cited in AI Overviews and AI Mode. On 11 November 2026 we checked every Google statement against current Google Search Central documentation (the AI features page, last updated 10 December 2025; the generative AI optimisation guide, last updated 10 July 2026; the robots meta tag specification, last updated 24 March 2026), Search Console Help, the Search Central blog and the Google Search Status Dashboard. Third-party studies are cited by publisher and date and use their own methods, mostly with US data. This is general information: Google changes these features often, and no vendor paid for inclusion. Researched with AI assistance and reviewed by our editorial team.
What Google has actually said
Google publishes two primary pages on this: AI features and your website, last updated 10 December 2025, and a longer guide to optimising for generative AI features, last updated 10 July 2026. Read both before you buy any AI Overview optimisation service. Their main points, with the Search Console help page for the control and the spam policies (last updated 28 August 2026), are summarised below.
| What Google says | Where |
|---|---|
| Nothing extra is required to appear in AI Overviews or AI Mode, and no special optimisation is needed | AI features page |
| A page must be indexed and eligible to show in Google Search with a snippet | AI features page |
| You do not need new machine-readable files, AI text files or extra markup | AI features page |
| There is no special schema.org structured data to add | AI features page |
| The site must be included in Search generative AI features in Search Console | Generative AI optimisation guide |
| Optimising for generative AI search is optimising for search, so it is still SEO | Generative AI optimisation guide |
| Google Search ignores LLMS.txt and similar files | Generative AI optimisation guide |
| No third-party tool can see Google's internal ranking or AI systems | Generative AI optimisation guide |
| Trying to manipulate generative AI responses in Search counts as spam | Spam policies |
| The exclusion control is not a ranking or inclusion signal for the rest of Search | Search generative AI control help |
The inclusion requirement in the fifth row is not an opt-in. Search Console's help says including your site is the default for all properties, and a property that has a parent inherits the parent's choice by default. You only need to act if someone has switched a property to exclude.
This guide is about earning citations. If you want to keep content out of AI answers instead, our guide to the UK page-level AI Overviews opt-out covers that decision.
How AI Overviews and AI Mode choose pages
Google says its generative AI features are rooted in its core Search ranking and quality systems, and names two techniques.
Grounding. Google describes retrieval-augmented generation: its core ranking systems retrieve relevant, current pages from the Search index, the model reviews them, and the answer shows clickable links to the pages that support it. A page outside the index, or one barred from snippets, cannot be a supporting link.
Query fan-out. The model issues several related searches at once to fetch more results. Google's example is a question about fixing a lawn full of weeds, which may fan out into searches about lawn herbicides, removing weeds without chemicals and preventing weeds. A page that answers one of those sub-questions well can be cited even if it does not rank for the original query.
Two more details matter. Google shows AI Overviews only when its systems judge that they add something to classic results, so they often do not appear. And AI Mode and AI Overviews may use different models and techniques, so their answers and links vary. Expect different citations for the same topic across the two.
Pew Research Center analysed the March 2025 browsing of 900 US adults (68,879 Google searches) and found that 18% of searches produced an AI summary. The share rose to 53% for searches of 10 words or more and to 60% for queries starting with question words such as who, what, when or why. Specific, question-led queries are where citations happen.
Step-by-step: how to get a page cited in AI Overviews
Step 1: Pass the eligibility checks
Run these checks on every page you want cited. Most never-cited problems end here. The requirements come from Google's AI features page, the robots meta tag specification and the Search Console control help.
| Check | How to verify | Fix if it fails |
|---|---|---|
| Page is indexed | URL Inspection tool in Search Console | Resolve the reason shown; list the URL in a sitemap |
| Page can show a snippet | No nosnippet or max-snippet:0 in the robots meta tag or X-Robots-Tag header | Remove the rule or raise the limit |
| Site is included in AI features | Search Console, Settings, Search generative AI | Choose include; check child URL-prefix properties too |
| Googlebot reaches the page | Crawl Stats report; server and CDN logs | Allow verified Googlebot through robots.txt, CDN and firewall rules |
| Main content is text | Rendered HTML in URL Inspection | Put key facts in HTML text, not only in images or script-loaded widgets |
A child property can override its parent. By default a property inherits the control from its closest parent that has changed it, so a forgotten /blog/ URL-prefix property set to exclude keeps that section out even when the domain is included.
Step 2: Get the ranking foundations right
Google's advice for AI features is to apply the same foundational SEO practice as for Search overall: meet the technical requirements, follow its policies and write helpful, people-first content. If those are weak, start with our step-by-step plan for ranking a website on Google's first page, then come back here.
Step 3: Write non-commodity content with a point of view
Google says content people find unique, compelling and useful will likely do more for visibility in generative AI search over time than any other suggestion in its guide. It contrasts commodity content, such as a generic list of tips for first-time homebuyers, with first-hand accounts that go beyond common knowledge, and asks writers not to repeat what is already online or what an AI model could easily produce.
In practice, publish your own numbers, process detail, worked examples, checked prices, and decisions you made and why. An AI answer that summarises ten pages needs a reason to link to yours.
Step 4: Answer the sub-questions inside the page
Map the questions a reader asks next and answer them in clearly headed sections on the same page. You do not need to cut text into tiny blocks: Google says its systems understand several topics on one page and there is no ideal page length.
Do not spin out a page for every fan-out variation. Google warns that creating separate content for each way people might search, mainly to manipulate rankings or AI responses, breaks its scaled content abuse policy.
Step 5: Keep key facts in text, and add images and video
Google lists important content in text form as a baseline, supported by high-quality images and video where they help. Video matters: in Ahrefs' March 2026 study, YouTube URLs made up 18.2% of cited pages that did not rank in the top 100 for the query, and 5.6% of all AI Overview URLs it analysed. If a topic is best shown, publish a video with an accurate title and description and embed it on the page.
Step 6: Keep structured data, Merchant Center and Business Profile accurate
Structured data is optional for AI features, but Google asks that it match the visible text on the page and that Merchant Center and Business Profile details stay current. Google says AI responses can include product listings and local business information, and that Merchant Center feeds and Business Profiles help products and services appear there.
Step 7: Provide a good page experience
Google's guide lists page experience among the practices that still matter: pages that display well on all devices, low latency, and main content that is easy to tell apart from everything else on the page. The Core Web Vitals targets are Largest Contentful Paint within 2.5 seconds, Interaction to Next Paint under 200 milliseconds and Cumulative Layout Shift under 0.1.
Step 8: Avoid shortcuts that count as spam
Google says chasing inauthentic mentions across the web is less useful than it looks, because its AI features depend on both its quality systems and its spam systems. Its spam policies now name attempts to manipulate generative AI responses. Paid listicle placements, fake reviews and mass-produced comparison pages carry the same risks as before.
Step 9: Measure, compare and iterate
Compare pages that earn AI feature impressions with pages that do not, looking at originality, depth on the sub-question and freshness of facts. Revisit quarterly rather than once: Ahrefs describes AI Overviews as probabilistic, with citations that shift between runs of the same query.
Time comparisons around ranking updates. Google released the November 2026 core update on 10 November 2026 and said the rollout may take up to two weeks, so compare before and after periods only once it completes. Our core update recovery playbook explains how to tell a ranking drop from clicks lost to AI answers.
What you can ignore, according to Google
Google's generative AI guide has a mythbusting section. For Google Search you can skip:
- LLMS.txt and special markup. Google Search does not use them, and publishing them for other services neither harms nor helps your Google visibility.
- Chunking content. Google's systems can handle several topics on one page and show the relevant part.
- Rewriting just for AI. AI systems understand synonyms, so you need not chase every long-tail phrasing.
- Inauthentic mentions. Core systems reward quality content and other systems block spam; AI features rely on both.
- Overfocusing on structured data. It is not required for AI features, though it still helps with rich results.
Google also warns against third-party tools that promise ranking success or claim to use internal Google metrics. AI visibility trackers sample prompts, so treat their numbers as estimates.
Snippet controls: nosnippet, max-snippet and data-nosnippet
AI is part of Search, so the controls are Search controls. Google says robots.txt rules for Googlebot govern crawling, while nosnippet, data-nosnippet, max-snippet and noindex limit what is shown. The table draws on the robots meta tag specification, the Search Console control help and Google's crawler documentation.
| Control | Scope | Effect on AI Overviews and AI Mode | Effect on classic results |
|---|---|---|---|
| nosnippet | Page | Content cannot be a direct input | No text snippet or video preview anywhere in Search |
| max-snippet:[n] | Page | Limits how much text can be a direct input | Snippet capped at n characters; 0 equals nosnippet; -1 lets Google choose the length |
| data-nosnippet | span, div or section element | Marked text is kept out of snippets | Same |
| noindex | Page | Page cannot appear | Page removed from Search |
| Search generative AI control | Domain or URL-prefix property | Links and content excluded, usually within 1 to 2 days | None, per Google |
| Google-Extended | robots.txt token | Not an exclusion; Search Console help points to it for limiting training of the models behind Search AI features | None; Google says it is not a Search inclusion or ranking signal |
The specification adds one caveat to max-snippet: the limit does not apply where a publisher has separately granted permission, for example through in-page structured data or a licence agreement with Google.
A typical pattern that keeps a page citable while protecting a proprietary table:
<meta name="robots" content="max-snippet:300">
<p>Public summary that readers and AI answers can quote.</p>
<div data-nosnippet>
Proprietary pricing table you do not want quoted.
</div>
If a preview control does not seem to work, use URL Inspection to confirm Googlebot sees it, then wait: Google says recrawling can take from several days to several months.
Two changes are coming. In the UK, the Competition and Markets Authority gave Google nine months from 3 June 2026 to give publishers effective controls over the use of their content in AI features such as AI Overviews, which takes the deadline to 3 March 2027. And on 30 October 2026 Google said it plans to support a shared AI training opt-out in robots.txt, based on IETF AIPREF work, in 2027; if a site uses it and does not name Google-Extended, Google plans to treat it as a Google-Extended disallow.
How to measure AI Overview and AI Mode visibility in Search Console
Performance report. AI Overview and AI Mode appearances count in the Web search type of the Performance report. An AI Overview occupies one position and all its links share it; a click on an external link in either feature counts as a click; a follow-up question in AI Mode counts as a new query; Search Labs experiments are excluded.
Generative AI performance report. Launched for some sites on 3 June 2026 and for all sites on 31 August 2026, the report shows impressions in AI Overviews and AI Mode, grouped by page, country, date or device. A search type filter separates Web: text-based from Web: multimodal, which covers Google Lens, Circle to Search on Android, uploaded images and Chrome image search; multimodal reporting arrived on 24 September 2026. The usual 1,000-row table limit applies, dates are in Pacific Time, and Discover has its own report. If the report is missing, your site either has too few AI impressions or has been excluded.
A simple monthly routine:
- Export the generative AI report by page and the Performance report by page for the same dates.
- Flag pages with AI impressions but falling classic clicks, and pages with classic rankings but no AI impressions.
- Check landing-page conversions in your analytics tool. Google says clicks from results pages with AI Overviews are higher quality, meaning visitors tend to spend more time on the site, so judge on outcomes as well as volume.
- Note each change you make and compare against untouched pages over at least a few weeks.
What third-party citation studies say
These studies use their own samples and methods, mostly with US data. Use them for direction, not targets. The figures below come from the two Ahrefs studies and Pew Research Center.
| Study | Date | Sample | Main finding |
|---|---|---|---|
| Ahrefs, first citation study | 21 July 2025 | 1.9 million citations from 1 million AI Overviews (top 3 citations in each) | 76.1% of cited pages ranked in the top 10; 9.5% in positions 11 to 100; 14.4% outside the top 100 |
| Ahrefs, update | 2 March 2026 | 863,000 keyword results pages, 4 million AI Overview URLs | 37.9% in the top 10 result blocks, 31.2% in blocks 11 to 100, 31.0% beyond 100 |
| Pew Research Center | 22 July 2025 | 68,879 Google searches by 900 US adults, March 2025 | Users clicked a traditional result on 8% of visits with an AI summary versus 15% without; 1% clicked a link inside the summary |
Two readings follow. Ranking still helps, because the top 10 remains the largest single band. But in the March 2026 data most cited pages sat outside it. Ahrefs suggests fan-out queries now drive more source selection, though it also improved its citation parsing between the two studies, so the fall from 76% to 38% is not a like-for-like comparison. Our data piece on whether ranking first still earns AI Overview citations compares more studies.
India notes for AI Overviews and AI Mode
Languages. AI Mode launched in India in English and Hindi, and on 8 October 2025 Google added Bengali, Kannada, Malayalam, Marathi, Tamil, Telugu and Urdu. If your buyers search in those languages, publish well-edited pages in them. Google's spam policies list mass-producing pages through automated transformations such as translation, without adding value, as scaled content abuse.
Voice and camera. Google says India is its largest user base for both voice and visual search. Check the multimodal filter in the generative AI report, and make product images clear, original and properly described.
Paid placements. On 28 October 2026 Google said it is scaling ads in AI Mode in India and nine other countries. Organic citations and ads are separate systems; our AI Mode ads playbook for India covers the paid side.
Regulated content. For health, finance and safety topics, Google says its systems give even more weight to content with strong E-E-A-T, and it encourages accurate authorship information. Named, qualified authors and clear sourcing are worth the effort on these pages.
For IT teams and publishers
- Bot access. Google lists allowing crawling in robots.txt and through any CDN or hosting layer as a baseline. Bot-management rules that challenge Googlebot remove you from AI answers and classic results together.
- Know which token does what. Blocking Google-Extended does not take your pages out of Google Search or AI Overviews; blocking Googlebot removes you from Search entirely.
- Template audits. A global
max-snippet:0ornosnippetin a CMS template, often added to protect paywalled pages, quietly removes AI eligibility across the site. Add it to release checks.
- Control governance. Property owners can change the Search generative AI setting, and top-level domain owners see a notice when a child property overrides theirs. Restrict owners and log changes.
- Evidence trail. Keep Search Console exports and server logs so you can explain a change in AI impressions after a release or a ranking update.
How eCorpIT can help
eCorpIT is a Gurugram-based technology organisation, founded in 2021, assessed at CMMI Level 5 and MSME certified, with senior-led engineering teams working across AWS, Microsoft and Google platforms. Our SEO, AEO and GEO audit service checks AI feature eligibility page by page, covering indexing, snippet rules, the Search generative AI setting, crawler access through your CDN and content gaps against the questions your buyers ask. Talk to us at /contact-us/.
Last updated: 11 November 2026.
