New

MCP: talk to your Google Search Console data and SEO analysis in Cursor, Claude, and other AI tools.

See MCP

Start here

Search all docs
Browse all topics

Canonical and robots meta possibilities

The canonical link element tells search engines which URL should be indexed when more than one address can show the same content. A missing canonical leaves that choice open. A canonical that points somewhere else asks search engines to prefer that other URL instead of the page you are looking at.

What we look for

  • No canonical URL — an indexable HTTP 200 page has no canonical link.
  • Canonical points somewhere else — the page has a canonical, and that href is not this page. A different path, host, or http versus https counts. A trailing slash does not: /products/leather-tote and /products/leather-tote/ are the same page. A relative href such as /products/leather-tote is resolved against the page before we compare.
  • Snippet controls — nosnippet, a very small max-snippet, noarchive, or nocache in robots meta. Those can limit how much text is shown or cited.

A missing robots meta tag is not a possibility. With no tag, the default is index, follow. We also skip the “points somewhere else” check on noindex pages and on utility or account paths such as login, cart, privacy, and thank-you.

What to do

Point the canonical at the URL of the page itself when that page should be indexed. If another URL really is the one that should rank, link to that URL and do not expect this address to stay in the index. Remove snippet limits that you did not mean to set.

Example

On https://example.com/products/leather-tote, a canonical aimed at a different product drops this page: 

<!-- Before: search engines are asked to index a different URL -->

<link rel="canonical" href="https://example.com/products/tote" />

<!-- After: this page -->

<link rel="canonical" href="https://example.com/products/leather-tote" />

The same page with a trailing slash in the href is already fine. We do not open a possibility for that.

Technical details

Pages marked noindex skip on-page possibilities, including this one. The same check runs on the free public scan for the URL you submit. In the app, the possibility lists the URL the canonical resolves to. Google can still choose a different canonical than the one you declared; that separate check uses a Search Console sample and is described in URL Inspection index signals. See also On-page and technical signals and How we find possibilities.