Help with scans and reports

Why did my website scan fail?

Check that the exact URL opens as a public web page without signing in. AEO test accepts HTTP and HTTPS pages, not PDFs, private network addresses, password-protected pages or API responses. If you copied a link from an editor or preview, use the published page’s URL.

A “hostname could not be resolved” message means the scanner could not find the address through DNS. Check the spelling and whether a newly configured domain is publicly reachable. A timeout can be temporary; wait before trying again.

HTTP 403 or a bot challenge can mean the website blocks automated requests even when it opens in your browser. HTTP 404 can mean the page has moved or no longer exists. Use the final URL shown in your browser after redirects. The scanner follows up to 2 redirects per resource.

The scanner also limits page size and complexity. The HTML response limit is 256,000 bytes of decoded content; a page with many nested elements can hit a separate parsing limit. These are scanner limits, not proof that your website is broken. Try a smaller representative page or ask your website manager to inspect the reported limit.

If a crawler policy result needs investigation, use the crawler access troubleshooting guide.

Why is Performance unavailable?

Performance measurements come from Google PageSpeed Insights. A provider quota, error, timeout or response-size limit can prevent us from retrieving them while the other checks still succeed. Open the unavailable finding to see the reason recorded for that scan.

Unavailable does not mean failed. Missing measurements are excluded from scoring. Continue with the findings that have evidence, and try a new scan later if you need fresh performance measurements. Reloading a saved report shows its original results; it does not rerun PageSpeed.

The performance tab describes a simulated mobile visit. Text can be the largest visible content, so a page without a large image can still have slow Largest Contentful Paint. Fonts, styles, scripts and server response time are also worth investigating. See the guide to PageSpeed results and slow LCP for choosing an improvement based on the measured delay.

What is the free scan limit?

The limit is 10 scans per client IP in a 2-hour window. Failed scan attempts also count once admitted. People using the same public network address may share this allowance. Wait for the window to reset before starting another scan. When a wait time is available, the limit message shows when you can retry in your local time. Select Analyze yourself when you are ready; the app does not retry automatically.

If the scanner is busy, wait 15 seconds and select Analyze again. A busy response does not use your scan allowance. This can happen when other visitors are scanning, including the same website.

Opening a saved report or using “Retry loading report” does not use a scan. Clearing browser history does not reset the allowance. You do not need an account or payment to use the tool.

Why will my saved report not open?

If loading times out or a connection fails, use “Retry loading report” to request the same saved result again. You can also reload the page. This does not start a new analysis.

Reports expire 30 days after the scan. An expired or incorrect link cannot retrieve the original report. Recent reports is a list stored in your browser, so an entry can remain there after its shared report expires. The list shows the shared link’s expiry, and the comparison picker labels expired links. These labels use your device’s time when the list opens; they do not check the server or remove your saved addresses. When the saved URL is available locally, we put it back in the input so you can choose to run a new scan.

Use Export → Download action plan for a text checklist you can send to your website manager. It includes suggested fixes in priority order, their evidence and instructions, plus a separate list of unavailable checks to verify. This download uses your saved results without running another scan.

Download JSON or use Print / Save PDF for the full report while it is still available. Downloaded files remain on your device after the shared link expires. Anyone with a shared report link can view it before expiry. See the privacy notice for how reports and local history are stored.

Why do results differ from what I see?

If your question is whether Google has indexed the page, use the search indexing troubleshooting guide. It explains which settings to inspect and how Search Console’s stored results differ from a live test.

The content checks read the HTML returned by the website without running its JavaScript. Your browser may load more text afterwards, or the server may return different content to an automated request. The scan covers one page and its supporting files, not the whole site.

Redirects can also differ. Cloudflare can skip automatic HTTPS redirects for scanner requests even when visitors are redirected. If Secure connection is unavailable after an HTTP scan, ask your website manager to verify the server redirect outside this scanner and scan the intended HTTPS address directly. This unavailable finding does not lower the score.

Compare the exact URL, scan date, scanner version, unavailable checks and performance measurement time before comparing scores. Changes in scoring rules or a provider’s measurements can affect results even when your page has not changed. A score is evidence about observable signals, not proof of search rankings or AI citations.

Use Compare saved reports to see changes in status and evidence side by side. Choose two reports from this browser or paste their shared links. This reads saved results and does not use your scan allowance. Anyone with the comparison link can view both reports until they expire.

Informational findings describe context or choices that do not lower the score. For example, restricting model-training access is a separate decision from allowing search crawlers. Read how scoring works before changing an intentional policy to satisfy a check.

Still need help?

Send the public page URL or report link, the exact error or check name, and what you expected to happen. Include the approximate time for a failed scan. Do not send passwords, access tokens or private links. Our Contact page has the support address and guidance for reporting an incorrect finding.