rel canonical · Canonical URL
The canonical tag indicates the reference URL among several versions of the same content. It consolidates signals (links, ranking) on a single page. Google treats it as a hint rather than a directive: it may choose a different canonical URL.
Standalone definition · 40 words
How does the canonical tag work?
Placed in the head (link rel=”canonical”) or sent in the HTTP header, it designates the preferred URL. Google groups duplicate or very similar pages, then picks a canonical URL from several signals: the tag, redirects, sitemap, internal links. According to the documentation, the tag is one signal among others.
What are the common mistakes?
Canonical pointing to a redirected or noindexed page, chains of canonicals, a relative tag resolved incorrectly, every paginated page canonicalized to page one, or a canonical that contradicts internal links and the sitemap. A contradictory signal leaves the choice to Google.
Canonical and GEO
An AI engine that retrieves several versions of the same page dilutes its chances of citing the intended URL. A consistent canonical, backed by redirects and internal linking to the reference URL, limits this risk. It does not replace a 301 redirect when a version must disappear.
Frequently asked questions
Canonical or 301?
A 301 removes the old URL for the user and the bot. A canonical leaves both versions accessible and designates the preferred one.
Is a self-referencing canonical useful?
Yes: it states the preferred URL against parameter variants.
Does Google always respect the canonical?
No. It is a signal; Google may pick another if it judges that signals contradict it.
Related terms
Further reading
- Search Central Live Deep Dive Barcelona 2026: what Google said over the 3 days
- Generative AI Report in Google Search Console: How to Analyze It
- Google Search Central Live Paris recap: Google Is Betting big on Agentic Commerce
- ← Back to the GEO glossary (114 terms)