The canonical link points to a page that does not work
The canonical link points to a URL that does not work: it answers 404, 410, 500, 501, 502 or 504, another 4xx code apart from 403 and 429, or the connection fails. Search engines are asked to index a page that is not there. Point the canonical to a working URL, usually this page itself.
What it means
The canonical https://example.com/a answers 403. Search engines may drop both pages. Point it to a working URL, usually this page itself.
Why it matters for AI search
A canonical tells search engines which URL to index. If that URL is gone or failing, they cannot follow the hint: they may ignore it or drop the pages from the index, and the page then loses its place in search and in AI answers that rely on search indexes.
The check fetches the canonical URL as a browser would. 403, 429 and 503 answers and timeouts are not counted, because a firewall or rate limit may be answering the check rather than the page being broken. Typical causes are a canonical that still names an old URL after a move, a typo in a template, or a trailing slash the server does not accept.
How to fix it
- Open the canonical URL from the finding in a browser and note what it returns.
- If the page moved, update the canonical to the new address. If the canonical was meant to be this page, use this page's own URL.
- Fix the template that builds the URL, not only this page, when the base URL, path or trailing slash is wrong everywhere.
- Run "Check again" on the page report to confirm the result.
Example
<link rel="canonical" href="https://example.com/pricing/"> How to confirm the fix
After the change, press "Check again" on the report: it lists what was fixed and how the score changed.