Key takeaways
- Absence from AI Overviews falls into three buckets: no overview appears, your page is not eligible, or your page is eligible but not the best answer.
- Google requires no special markup for AI features. Standard indexability, snippet eligibility, and page quality are what matter.
- Query fan-out means Google may pull sub-answers from multiple sources. Your page needs to win specific sub-questions, not just the general topic.
- Track visibility by prompt family over time, not by a single query, and re-check 4 to 6 weeks after making changes.
Does Google show an AI Overview for the query at all?
Start by checking whether Google shows an AI Overview for the query, because you cannot be cited in an overview that does not appear. Many searches, especially highly local, transactional, or YMYL-adjacent ones, still return classic results with no generative summary at all.
Before assuming a technical problem, confirm the basics. AI Overview display depends on the exact query wording, the device you search from, your location, whether you are logged in, and ongoing personalization that Google applies per session. A query that shows an overview on desktop in one city may show nothing on mobile in another, which has nothing to do with your website.
Keep a simple log when you test: the exact prompt text, device type, location used, date, and whether an overview appeared at all versus which sources it cited. This log matters because query behavior shifts over weeks as Google adjusts trigger rates. Without it, you cannot tell whether a later change fixed something or whether the query simply stopped triggering an overview that day.
Is your page eligible for Google AI features?
A page must be indexed and snippet-eligible before it can reliably appear as a supporting link in Google AI Overviews or AI Mode. Google has stated there are no additional technical requirements beyond standard Search and snippet eligibility to be considered.
Use URL Inspection in Search Console to confirm the page is actually indexed and that the live version matches what you expect Google to see. Check for accidental noindex tags, blocked robots.txt paths, canonical tags pointing elsewhere, or meta directives that suppress snippets. If Google cannot show a normal snippet for your page in classic search, it has no reason to be able to use it inside an AI Overview either.
JavaScript-rendered content is a common silent blocker. If your key facts, pricing, hours, or service details only render client-side after heavy scripting, crawlers may see an empty or thin page. Google's generative AI guidance reiterates that crawlability, technical clarity, and avoiding duplicate content remain foundational, and no special AI markup changes that baseline requirement.
Is your content the best support for the answer?
Eligibility is not enough. Your page also needs to answer the specific sub-question Google's system is trying to satisfy for that query, which often means matching a narrower intent than your homepage or general service page covers.
Google's AI features can use query fan-out, breaking one search into several underlying questions and pulling a wider, more diverse set of sources than a single classic result page would show. A page that is technically indexed but only addresses the topic in general terms will lose to a page that answers the exact sub-question directly and with specifics.
Consider 'best emergency plumber near Austin.' The fan-out likely includes location coverage, true emergency availability, response hours, licensing or insurance proof, and review signals. A generic plumbing services page mentioning Austin once will rarely beat a page that explicitly states service area, 24-hour dispatch, credentials, and recent reviews in structured, scannable language.
Are third-party sources saying more than your site does?
If directories, review sites, Reddit threads, or industry roundup lists explain your category, comparison, or reputation better than your own site does, Google may cite them instead of you, even when your business is well known offline.
This happens constantly in local and comparison-style queries. A third-party review aggregator often states pricing ranges, service comparisons, and pros and cons more plainly than a business's own marketing copy, which tends to avoid specifics or comparative language for brand reasons. Google's systems favor whichever source most directly and confidently answers the sub-question being asked.
This is a source gap, not a penalty. It means the information gap between what your site says and what the fan-out question needs is being filled somewhere else. Auditing which sources currently outrank your own pages for these sub-questions tells you exactly what content or clarity is missing, and where competitors or third parties are winning the citation instead.
What should you fix first?
Fix the highest-confidence blockers first: indexability, snippet eligibility, page specificity for the exact sub-question, closing source gaps versus third-party citations, and correcting outdated local or directory facts that contradict your own site.
Start with a short prioritized pass. Confirm indexing and snippet eligibility using URL Inspection, since nothing else matters if the page is blocked. Next, check whether the page answers the specific sub-question a user is likely asking rather than a broad topic. Then compare your content against whichever third-party source currently gets cited for related queries, and note what facts or framing they include that you do not.
Finally, audit outdated business listings, directory profiles, or old pages with stale hours, pricing, or service areas. Conflicting facts across the web reduce Google's confidence in any single source, including your own site. Structured data such as FAQ schema can support clear, machine-readable answers, but Google is explicit that structured data enables eligibility for rich results without guaranteeing they will display.
How should you measure progress?
Measure progress by prompt family, not by one query. Track whether you appear, whether you are cited, which source gets cited instead, and whether the underlying answer reasoning shifts toward your content over repeated checks.
A single query is noisy and can flip for reasons unrelated to your site, so build a small set of related prompts covering your core services, locations, and comparison questions. Re-run that set on a schedule, log the results, and look for directional trends rather than one-off wins or losses. This is the same logic behind consistent AI visibility scoring rather than one-time spot checks.
After making changes, especially indexing fixes or new content, allow roughly 4 to 6 weeks and confirm through Search Console that pages were recrawled and reindexed before re-testing prompts. Changing content and immediately re-checking the same day usually just measures crawl delay, not whether the fix actually worked. Patience paired with a repeatable prompt log is what separates real progress from noise.
Reader questions
Frequently asked questions
Can I force Google to show an AI Overview?
No. Google decides per query, per session, and per user whether to show an AI Overview at all, based on factors you do not control. You can only make your content eligible and well suited to being used if an overview does appear.
Does schema make my site appear in AI Overviews?
No. Structured data can make your page eligible for rich results and clearer machine parsing, but Google is explicit that it does not guarantee display in AI Overviews or anywhere else. It supports clarity, it does not force inclusion.
Why does Google cite Reddit instead of my website?
Reddit threads often answer specific sub-questions more directly, with comparative language and firsthand detail that business sites avoid. If a thread states pros, cons, or pricing plainly where your site stays vague, it can be the stronger match for that sub-question.
Should I create pages for every AI Overview prompt?
No. Building a separate page for every slight query variation creates thin, duplicate content that hurts overall quality. Instead, build fewer, stronger pages that clearly answer the real sub-questions behind a group of related prompts.
How often should I check AI Overview visibility?
Check on a recurring schedule, such as every few weeks, using the same prompt set each time rather than one-off checks. Allow 4 to 6 weeks after any content or indexing fix before expecting a measurable shift in citations.