A canonical URL tells search engines which version of a page is the "master" copy when multiple URLs show the same or very similar content — preventing duplicate content from splitting ranking signals across near-identical pages.
Common situations that need one
- The same page reachable via both
http://andhttps://, or with and withoutwww - URL parameters that don’t change the core content (tracking parameters, sort order, session IDs)
- Product pages accessible via multiple category paths in an e-commerce site
How it’s implemented
A <link rel="canonical" href="..."> tag in the page’s <head>, pointing to the preferred URL — even a page pointing to itself as canonical is standard, correct practice.
FAQ
Does every single page need an explicit canonical tag?
It’s good practice to self-reference on every page, since it removes ambiguity even in cases you didn’t anticipate (like an unexpected URL parameter being crawled).