Answer first: URL Inspection tells you two different things about one URL on a Search Console property you verified. The first reading is Google’s indexed copy: whether that URL is eligible to appear in Google Search. The second is a live test: whether Google-InspectionTool can fetch the page now. “URL is on Google” means eligible, not that the page is showing, and not that it is a supporting link in an AI Overview or AI Mode. Google’s AI features guide says a supporting link must be indexed and eligible for a snippet. This tool is how you check the index half of that sentence, and how you read the HTML from the last crawl. It does not list queries, impressions, or citations. Recheck the help pages below before you brief anyone from a status line. This article records them for a reader on 16 October 2026.

What are the two readings?

Open the verified property and use the URL Inspection tool. Type the full URL into the bar at the top, or use Inspect beside a URL in another report. The URL has to belong to the property you have open. A .im host and a .co.uk host are two inspections. A staging hostname is not a reading of the public site.

The first screen is the indexed reading. Google says it is not a live test. It describes the most recently indexed version, which is the version used in Search. The page on the server may have changed since that crawl. Last crawl is shown in your local time, and every field on that screen comes from that crawl. Do not copy the Pacific Time rule from the Generative AI performance report onto this timestamp. That rule belongs to the report’s dates, not to this tool.

The live test is the button after that. Google says it fetches the URL in real time, and that a valid result only confirms Google-InspectionTool can access the page for indexing. There is no test that guarantees the page will enter the index. You can switch between Google Index and Live Test. They are allowed to disagree. A fix you made this morning can look fine on the live test while the indexed reading still describes yesterday’s page.

There is a daily limit on inspection requests for each property, and a separate per-property daily limit on live tests. Google’s help page, in the text recorded here, does not publish either number. Do not invent one, and do not keep re-running the test to force a different sentence.

What does “URL is on Google” mean?

Read the line at the top before you read anything else.

  • URL is on Google. The URL has been indexed and can appear in Google Search, and Google found no problems with the enhancements it detected, such as structured data or a linked AMP page. Google says this does not guarantee that the page is appearing. The tool does not check every condition. The check Google names, for whether the URL is appearing at all, is to search for that URL on Google.
  • URL is on Google, but has issues. The same indexed state, plus a problem that might stop an enhancement. Google’s examples are an AMP problem, or malformed structured data for a rich result such as a recipe or a job posting. This is not the same sentence as “the page cannot be a supporting link”.
  • URL is not on Google. This URL will not appear in Google Search results. The reason is in the Page indexing section. Decide whether that reason is one you meant.
  • URL is an alternate version. The URL is an AMP alternate of a canonical non-AMP page. Read the Google-selected canonical and check that it is the URL you expect.

A search for the URL answers a narrower question than an AI citation. It tells you whether that URL is appearing as itself. It does not tell you whether an AI Overview used the page, and it does not list the query. Impressions inside AI Overviews and AI Mode are the performance report. If the URL is on Google in this tool and the report is missing, those are different facts. Google’s reasons for a missing report are not enough impressions, or exclusion. They are not a second copy of this status line.

If the tool says the URL is on Google and a search for the URL shows nothing, Google’s troubleshooting steps say to check the Manual Actions report, the Security Issues report, and whether someone temporarily blocked the page. This tool does not test those. A clear inspection is not a clearance certificate.

Which fields matter before anyone talks about a citation?

Google’s technical requirements say a page is eligible to be indexed when three things are true: Googlebot is not blocked, Google receives an HTTP 200 success status, and the page has indexable content. Indexable content means the text is in a file type Google Search supports, and the content does not violate spam policies. Google says meeting those requirements still does not mean the page will be indexed.

URL Inspection is the one-URL place Google names for that check. The Page indexing section is the part to read. For a small firm, these fields are the ones that change the next action:

Crawl allowed? This is the robots.txt answer. If it is No, Google cannot fetch the page, and it cannot read a meta tag on it. Google says Indexing allowed will still say Yes in that case, because a noindex tag on a blocked URL is invisible. The page might still appear in Search results. Do not celebrate “Indexing allowed: Yes” while crawl is blocked. Remove the robots.txt block if you want the page fetched, then test again. Which token is Googlebot, and how that differs from Google-Extended, is the citations guide. This screen does not show Google-Extended.

Page fetch. Crawl can be allowed and the fetch can still fail. Google’s documented failures include a server it could not reach, an invalid SSL certificate (Google will not test an HTTPS URL on the site unless the certificate is valid), an unreachable robots.txt (Google will not crawl the site if that file exists but cannot be fetched), a DNS answer that points at a private address, and hostload exceeded. Hostload means Google thinks the site is already at capacity for crawling or inspection, and that tests wait until that load drops. Stop clicking Live test if the field says that. Read the reason. Do not buy a citation product because the host returned an error.

Indexing allowed? This is whether the page itself disallows indexing, usually with noindex. If it says No, the page will not appear in Google Search until that tag or header is removed. If indexifembedded is present as well as noindex, Google says the page is indexed only when it is embedded. A theme can add noindex without anyone choosing it. The preview rules that decide snippet text, including nosnippet and max-snippet, are a different layer. They are written up in the include or exclude guide. Inspection tells you whether the crawled HTML contains the tag. It does not replace that guide.

User-declared canonical and Google-selected canonical. The declared canonical is the one your page names. None is a normal value when the page is not one of a set of alternates. The Google-selected canonical is the URL Google treated as the authoritative copy. Google may pick yours, or it may pick another. You can see that choice only on the indexed reading. Google says the live test cannot predict it, and that the value can be a few hours behind the index. The canonical is also not always the URL a searcher sees: a desktop URL and a mobile URL can both exist, and Google says it will probably show the one that matches the device.

If you inspect a URL that redirects, the indexed reading describes the URL you typed, not the target. Google says to press Inspect on the canonical shown under Page indexing when you want the target’s index status. The live test does the opposite of what people expect: it follows the redirect, then tests the destination, and it does not say that it followed one or which URL it tested. A firm that moved /contact/ to /quote/ can stare at a green live test and think it checked the old address. Read the indexed redirect note, then inspect the URL you actually want understood.

Sitemaps and referring page. A sitemap row lists sitemaps submitted in the Sitemaps report, or named in robots.txt, that point at this URL. “No referring sitemap” means Google could not identify one. Google says sitemaps found some other way are not listed, and that there is a known gap where a submitted sitemap is sometimes not reported. The technical requirements do not include “has a sitemap” as one of the three conditions. A blank sitemap row is not, by itself, why an AI Overview is missing. The referring page is a URL Google may have used to discover this one, and it might be a parent of a parent. If the field is empty, Google says that does not prove no referring page exists. Keep the internal link anyway. The AI features guide lists internal links among the ordinary SEO habits that still apply.

Crawled as, and the HTML. The indexed reading names the user-agent type, desktop or mobile. View crawled page shows the HTTP response and the HTML, when the status is “URL is on Google” or “URL is on Google, but has issues”. If that control is disabled, Google says to hover it for the reason. The live test has its own View tested page, and a screenshot, only when the live fetch succeeded. Google says the screenshot is how Google-InspectionTool sees the page, and that the screenshot and the details are taken from the mobile version. A price that exists only on a desktop layout is not what that screenshot is for.

The useful AEO check is the HTML, not the screenshot. Google’s troubleshooting note for preview controls says to use this tool to see the HTML Googlebot received while crawling. Confirm the sentence you want understood is in that HTML as text. A picture of a van, a slide, or a rate card is a weak place for the only clear fact. If the URL is the quote form, labels and a real submit control are a separate job, covered in the quote form guide. This tool will not grade the form.

What does the live test refuse to decide?

Google says a positive live result means the URL can be crawled and parsed. It does not mean the page will be indexed, and it does not mean the page conforms to quality or security guidelines. The live test also skips manual actions, legal removals, and URLs blocked with the removal tool.

It does not look for the URL in a sitemap or on a referring page. It does not decide duplicates. Google says duplicate pages are not indexed, and that canonical selection happens at indexing time. The live test can say “URL is available to Google” while the index still says crawled but not indexed, or discovered but not indexed, or that Google chose a different canonical. Those states are on Google’s list of index conditions the live test does not check. “Page indexed without content” is on that list as well. A green live test is not proof the indexed copy contains your paragraph.

So the order is fixed. Read the indexed status first. Use the live test only to see whether a change you have already made is visible to a fetch now. Then wait for a new crawl before you treat the indexed line as updated.

When is Request indexing worth using?

Request indexing asks Google to crawl that URL. Google says indexing can take up to a week or two, and also that it typically takes a day or so but can take much longer. Those are both on the same help page. Do not average them into a promise, and do not treat a quiet week as a broken button.

Google also says the request does not guarantee that the page will appear in the index. You cannot send the request if the live test already calls the page non-indexable. There is a daily limit on how many requests you can submit. The number is not on the page recorded here. If many URLs changed, Google’s own suggestion is a sitemap that marks the updated pages, not a morning spent clicking one URL at a time.

Use the button after a real change: you removed a noindex you did not mean, you fixed a redirect, or you replaced a wrong phone number in the HTML. Asking again because an AI chat did not mention you is not a change to the page. If the status is “URL is unknown to Google”, Google’s troubleshooting note says the URL has not been seen, and that requesting indexing typically takes a few days. That sentence is about an unknown URL. It is not a forecast for AI Overviews.

For a preview-control change, such as adding or removing nosnippet, the AI features guide gives a wider range: crawling can take from several days to several months, depending on how often Google decides the page needs fetching. You can request a recrawl after a real change. Judge the result against that note, not against a search from the same hour. The control guide is where those tags are explained. This article only tells you how to see whether the crawled HTML matches the tag you think you published.

How does this sit next to an AI Overview?

Google’s AI features guide says that to be eligible as a supporting link in AI Overviews or AI Mode, a page must be indexed and eligible to be shown in Google Search with a snippet, and it must meet the ordinary technical requirements. There are no additional technical requirements. There is no special schema to add, and no machine-readable AI file you have to publish. Google also says that meeting the requirements, the best practices, and the policies still does not mean Google will crawl, index, or serve the page.

Read that as a gate, not a prize. URL Inspection can fail the gate in public: not indexed, blocked, noindex, or a canonical you did not intend. Passing it means the URL is eligible to appear in Search, which is what “URL is on Google” already says. It is not a citation.

Two neighbouring controls are easy to mix up with this screen.

The Search Console setting Include, Exclude, or Inherit decides whether the property may be used in Search generative AI features. An indexed URL on an excluded property is still indexed. Exclusion does not rewrite this tool’s status into “URL is not on Google”. That setting is the control guide. Snippet eligibility is the other half of Google’s sentence, and it lives in the meta tags that guide explains. A page can be on Google and still be nosnippet.

ChatGPT, Perplexity, Bing, Copilot, and Apple do not read this screen. Their crawler choices are the citations guide and the Perplexity, Bing, Copilot and Apple guide. A pass here does not travel.

Google also says you do not need a special schema for these AI features, and that structured data should match the visible text. An enhancement warning can leave the URL on Google with issues. That warning is about the rich result, not an automatic block on a supporting link. Fix markup that disagrees with the page, because a wrong phone number in the markup is still a wrong phone number. Do not add FAQ markup because you want an overview. How to correct a fact an assistant is already repeating is the wrong business details guide. How to record where a published claim came from is the proof ledger guide.

What should a local firm still read on the page?

Inspection does not check that the coverage sentence is true. It checks whether Google could fetch a URL and what it selected as canonical.

Open the Google-selected canonical, logged out. The first screen should name the service and the places you actually cover, in text. A Douglas firm that only works on the Island should not leave a sentence that reads as UK-wide attendance. A firm that covers three towns should name those towns, not a near-copy page for every parish. The inspection status of those copies will not make the coverage honest. How to write that sentence is the local citations guide. How to shape the page around one question is the service-page guide.

If the crawled HTML still shows the old number, the old town, or last year’s price, the index is repeating the page Google fetched. Changing a social profile does not edit that HTML. Change the page, confirm the live test can see the new text, and only then request indexing.

Google’s technical page also says to use the Page indexing report and the Crawl stats report together when you want a list of URLs Google cannot reach. Each report can show different URLs. Inspection is for the one address you care about. It is not a substitute for those lists, and those lists are not a citation report.

What should I check on the day I open it?

Do this on the property customers use. Write the date next to the status line.

  1. Inspect the exact URL you want understood, including the host customers type.
  2. Read the indexed line first: on Google, on Google with issues, or not on Google.
  3. Read Crawl allowed, Page fetch, and Indexing allowed. Ignore a Yes on indexing if crawl is blocked.
  4. Read the Google-selected canonical. If it is a different URL, inspect that URL before you judge this one.
  5. If the indexed HTML is available, confirm the answer is text in that HTML. If you have changed the page since the last crawl, run one live test and read View tested page.
  6. If you need snippet text suppressed or restored, compare that HTML with the tag you intended, then use the control guide for the tag. Do not expect this tool to apply it today.
  7. Search for the URL on Google if you need Google’s own check of whether the URL is appearing. Use the performance report, not this tool, for AI Overview and AI Mode impressions.
  8. If the public page is the thing you want reviewed, the visibility audit is the page-and-evidence pass. It does not see inside the model, and neither does URL Inspection.

What should I not claim?

  • That “URL is on Google” is an AI Overview, a citation, a click, or an enquiry.
  • That a valid live test means the page is indexed, or that it will be.
  • That Request indexing has a clock you can put in a proposal. Google publishes more than one range, and a request is not a guarantee.
  • That a structured-data warning, or the absence of FAQ markup, is why an assistant did not name you. Google says there is no special schema for these AI features.
  • That this tool measures ChatGPT, Perplexity, Bing, Copilot, or Apple.
  • That a blank sitemap row or a blank referring-page row proves the page cannot be found.
  • A result, a town, or a before-and-after Halo has not published with evidence. This guide does not add one.

Where Halo fits

Halo’s search, AEO and content service covers the public page: the answer in the first screen, internal links, and structured data that matches the visible words. URL Inspection sits in the Search Console property you verify. Halo does not need that login. If you already have the screen, the useful note is the status line, Crawl allowed, Indexing allowed, the Google-selected canonical, and the last crawl date.

To start, send the URL through the free AI audit. Halo does not guarantee indexing, impressions, citations, or enquiries.

FAQ

Does “URL is on Google” mean my page will appear in an AI Overview?

No. Google says that status means the URL is indexed and can appear in Search, and that the tool does not check every condition for appearing. A supporting link in AI Overviews or AI Mode also has to be eligible for a snippet. Neither reading is a citation, a click, or an enquiry.

Is a successful live test the same as being indexed?

No. Google says the live test only confirms that Google-InspectionTool can access the page. A valid result does not guarantee indexing. Canonical selection and several not-indexed states are decided later, and the live test does not check them. The indexed reading is the one that describes the copy used in Search.

Why does Indexing allowed say Yes when robots.txt blocks the page?

Because Google cannot see a noindex tag on a URL it is not allowed to fetch. The help page says Indexing allowed stays Yes in that case, and that the page might still appear in Search results. Crawl allowed is the field that shows the robots.txt block.

Can the live test tell me which URL Google treats as canonical?

No. Google says the Google-selected canonical is available on the indexed reading only, and that the live test cannot predict it. The live test also follows redirects without telling you the final URL. Read the indexed canonical, then inspect that URL.

Does a structured-data warning in URL Inspection block AI Overviews?

Not by itself. “URL is on Google, but has issues” means the URL is still indexed, and that an enhancement such as structured data has a problem. Google’s AI features guide says there is no special schema required for AI Overviews or AI Mode. A warning on a rich result is not the same status as “URL is not on Google”.

Does an Isle of Man firm use a different URL Inspection tool?

No. The tool reads whichever verified property contains the URL. A .im host and a .co.uk host are separate properties. The help page recorded here does not describe an Isle of Man mode. The coverage sentence on the page still has to name where you actually work.

Sources

  • Google Search Console Help, URL Inspection tool, including indexed and live readings, the status lines, Page indexing fields, canonical selection, redirects, live-test limits, Google-InspectionTool, request indexing, and the conditions the live test does not check.
  • Google Search Central, AI features and your website, including supporting-link eligibility, the statement that indexing and serving are not guaranteed, the note that no special schema is required, and the URL Inspection step for preview controls.
  • Google Search Central, Google Search technical requirements, including the three eligibility conditions, HTTP 200, and the instruction to test a specific page with URL Inspection.