Canonical tags and AI Overviews

The canonical tag is a single line of HTML:

<link rel="canonical" href="https://example.com/blog/the-real-post-url">

It does two things that affect AI Overviews:

1. It consolidates ranking signals. Without a canonical, Google may split signal between https://example.com/blog/post, https://www.example.com/blog/post, https://example.com/blog/post?utm_source=x, and so on. Split signal = weaker page = less likely to be cited in AI Overviews.
2. It picks which URL gets cited. When Google does decide to cite your content, it cites the canonical URL. If your canonical points to an unindexable page (or to a competitor by mistake - this happens), the citation goes there.

We have seen real cases where a canonical tag was set up via a CMS plugin and pointed to the staging environment URL of the same article. The production article was indexed but the canonical tag pointed at a noindexed staging copy. AI Overviews cited nothing.

Two failure modes are common enough to call out:

Canonical to a noindex page. The signals consolidate to a page that cannot be indexed, so neither URL appears.

Canonical to a different domain. Usually accidental, sometimes the result of a content syndication setup. Citations go to the other domain.

The baseline

Every important page should have a self-referencing canonical:

<link rel="canonical" href="https://example.com/blog/post-slug">

The URL inside href should be:

• The full absolute URL (including https:// and the domain)
• The version you want indexed (not the variant with tracking parameters, not the trailing-slash version if your site uses no trailing slashes)
• A page that itself is indexable (no noindex, returns 200)

When you have duplicates

If the same content lives at multiple URLs, all of them should have a canonical pointing to the same chosen URL. That includes parameter variations.

<!-- Both of these point to the same canonical -->
<link rel="canonical" href="https://example.com/products/widget">

<!-- ...even on https://example.com/products/widget?ref=abc -->
<link rel="canonical" href="https://example.com/products/widget">

When you syndicate content

If your article is republished on another site (legitimate syndication), the syndicating site should set their canonical to your original URL:

<!-- On the syndicating partner's site -->
<link rel="canonical" href="https://example.com/blog/original-post">

This is how Google's syndication guidance recommends preserving the original site's ranking and citation signal.

Verify

Run Google's URL Inspection tool on the URL. Under "Indexing," it shows both your declared canonical and the canonical Google actually chose. If they differ, Google has decided to use a different URL, and you should investigate why (often a stronger duplicate signal, a redirect chain, or a sitemap mismatch).

• Canonical pointing to a noindex page. Both pages effectively drop out.
• Canonical pointing to a 404 or redirect. Google treats the canonical as broken and picks its own.
• Relative URL in the href. Always use absolute URLs (https://...) for canonicals.
• Different canonicals on different variants of the same page. Defeats the purpose. All duplicates should point to the same canonical.
• No canonical at all on parameterized pages. Each parameter variant gets indexed separately and dilutes the signal.
• Cross-domain canonical by accident. Easy to do when copying a CMS template. Always check the canonical's domain matches the page's domain unless you intend syndication.