Glossary /Technical SEO
Canonical Tag
A link element naming the preferred URL for a page, used to consolidate duplicates so ranking signals land on one address.
A canonical tag declares which URL is the real one when the same content is reachable at several addresses. Tracking parameters, session IDs, trailing-slash variants, HTTP and HTTPS, www and apex — each creates a distinct URL, and without a canonical the engine has to guess which to keep and may split the signals between them.
The tag is a strong hint, not a directive. An engine will override a canonical it does not believe, most often when the nominated page is substantially different from the one making the claim. Canonicals that point somewhere unrelated tend to be ignored entirely.
Two failure modes are common enough to check for by default: every page canonicalising to the homepage, which is what happens when a template hardcodes the value and it deindexes the whole site; and canonicals pointing at a redirecting URL, which forces the engine through a hop it did not need to take.