Back to blog

Not Showing Up in Google AI Overviews? A Website Checklist

Find out what to check when your site is missing from Google AI Overviews: indexing, Search Console controls, content gaps, and how to measure visibility.

Published:
Updated:
Reading time
5 min read

Check eligibility and content when your site is missing

A page can appear in ordinary Google results without being linked in an AI Overview. Rewriting it immediately can waste effort if a publishing control is excluding the site.

Check eligibility first, then improve the content. This guide is for website owners investigating missing links to their content. It does not cover troubleshooting an AI Overview that fails to appear in a searcher's interface.

The checklist and writing examples below are our suggested workflow, not Google's scoring criteria. Product details were checked against official documentation on September 7, 2026. Meeting the requirements does not guarantee a citation.

What makes a page eligible for Google AI Overviews?

Google says a page must be indexed and eligible to appear in Search with a snippet to qualify for its generative AI features. The site must also be included in those features through its Search Console controls. Eligibility does not guarantee that Google will show the page. See Google's guidance for generative AI search.

An indexed URL is therefore only part of the investigation. Confirm that the site's publishing settings also match its intended visibility.

Review the Search generative AI control

In Search Console, open Settings → Search generative AI to check whether your site's links and content are included or excluded. If the property inherits its setting, check the parent property's choice as well. Google says this control was rolled out to all websites worldwide as of August 31, 2026. See the Search generative AI control documentation.

For example, an article may have no local publishing restrictions but inherit an exclusion from its parent property. Adding more paragraphs will not resolve that exclusion. Confirm the intended publishing policy with the site owner before changing the control.

What to check when your site is missing

Work through the table from top to bottom. Each item is a possible issue to investigate, not a confirmed cause. Use the checklist to identify problems and work with the person responsible to address them.

Check Action
Can visitors read the public URL? Login requirements, error pages, and access restrictions
Can Googlebot fetch the content? robots.txt rules and server or CDN restrictions
Is the intended URL indexed? Its indexing status in Search Console
Are search previews restricted? noindex and snippet-related directives
Is the site included in generative AI search? Search Console controls and inherited settings
Does the page answer the intended question? A direct answer, conditions, and exceptions
Can the reader verify its claims? Measurements, comparison criteria, and primary sources
Is there a useful next step? Relevant guides or service details

Distinguish robots.txt from noindex

robots.txt primarily controls crawler access. It does not reliably remove a URL from search results. Google explains that a blocked URL may still appear when it is discovered through links elsewhere. See Google's introduction to robots.txt.

A URL appearing in Search is therefore not proof that Google can retrieve its body text. When asking a developer to investigate, share the symptoms and settings you checked alongside the URL.

Keep a short investigation record

Record the following for each page so editors and developers can work from the same evidence:

  • The exact URL and date checked.
  • Whether it appears in ordinary search results.
  • The indexing status shown in Search Console.
  • The generative AI control value and whether it is inherited.
  • Whether the content is readable without signing in.
  • Recent changes and their dates.

Mark anything you have not checked as unverified. Treating an assumption as a completed check can send the investigation back to the beginning.

If the settings are correct, make the answer more specific

For a content audit, write down four things from the page: question, answer, conditions, and evidence. Doing this before rewriting the article helps identify exactly what information is missing.

Element Content Example of a gap
Question The specific question the page resolves Only a broad topic has been chosen
Answer The conclusion the reader needs first Background continues without an answer
Conditions Where the answer applies and its exceptions The claim seems to apply to every situation
Evidence Sources or records that support the conclusion Evaluative words replace verifiable details

Example: make a service pricing page useful for comparison

Before

We build high-quality websites at affordable prices. Our extensive experience helps us recommend the perfect package for your business.

After

A five-page company website starts at JPY 330,000, including tax. Delivery typically takes six weeks from kickoff when you supply the copy and photography. The price includes design, development, and a contact form. Interviews, photography, and a booking system are quoted separately.

In this example, the revised version lets readers compare the price, scope, requirements, timeline, and additional costs. For a real page, use approved pricing and delivery terms, then link to the relevant service documentation.

These details help a buyer decide whether the offer fits their project. Evaluate the quality of resulting inquiries as well as any change in AI citations.

Do not start by adding llms.txt or structured data

Google says llms.txt and special structured data are not required for its generative AI search features. Adding them does not resolve missing eligibility requirements or unsupported claims. See the section on unnecessary tactics in Google's official guide.

If a service page omits what the price includes, publish that information first. When using structured data to describe an article or author, keep it consistent with the information readers can see.

For the broader content and technical foundations, read our guide to AI search optimization and creating citable content.

Measure visibility separately from business results

Search Console's generative AI performance report shows impressions from AI Overviews and AI Mode. It supports analysis by page, country, date, and device. A report may be unavailable when the site has not received enough impressions, among other reasons. See the official generative AI performance report documentation.

An impression is not an inquiry. Keep the following records separate so each metric answers the right question.

Record What it helps assess
The page's generative AI impressions Whether its visibility has changed
Ordinary search impressions and clicks Whether search performance is changing more broadly
Attributable visits and inquiries in analytics Whether visitors take a useful business action
Update dates and changed sections Which revisions preceded a change in performance

For example, compare the 28 days before a revision with the 28 days after it. This is a suggested reporting interval, not a claim that improvements take 28 days. Note seasonal demand, advertising changes, and other activity that could affect the comparison.

Also preserve the scope of the metric. Combined impressions from AI Overviews and AI Mode do not tell you how often a page was cited for a particular question. Avoid attributing every change to the article revision alone.

Start with one page that matters to the business

Choose a service page that supports inquiries or an existing article relevant to your offer. Address access and eligibility problems first, then add missing conditions and evidence to the content.

Please Cite Me! audits a public URL’s technical settings and content to show what needs improvement. It checks whether the page can be accessed, understood, and verified, with concrete guidance you can follow without prior experience in AI search optimization. Start with your page URL to see what to fix first.

About the author

Yuma Okamoto

Yuma Okamoto

Developer of Please Cite Me! and CEO of Abeam Tech Inc.

University of Tokyo graduate focused on web product development, SEO, and information architecture for AI search, with a particular interest in how web content is understood and cited by search engines and AI.

X / @okamoto_yuma_

Frequently asked questions

Does appearing in Google Search guarantee an AI Overview citation?

No. Meeting the eligibility requirements is different from being selected as a supporting source for an answer. Check indexing, snippet restrictions, and generative AI inclusion settings before reviewing the content.

Can I submit my page for inclusion in AI Overviews?

The official guidance referenced here does not provide an application that guarantees inclusion. Choosing to include a site in Search Console's generative AI features is an eligibility setting, not a request for a citation in a particular answer.

Will adding llms.txt make my site appear?

No. Google says llms.txt is not required for its generative AI search features. Check access, indexing, inclusion settings, and the information on the page first.

Please Cite Me!

Audit your page for AI search

Please Cite Me! reviews a public page's technical access and editorial clarity, then turns the evidence into a prioritized report.

Audit a page