Canonical tag: A tag in a page’s head that names the preferred address for its content when the same or very similar content is reachable at more than one URL. Google treats it as a strong signal, not an order.
When you need one
Duplicate URLs creep in on any WordPress site: tracking parameters, print versions, a product reachable under two categories in an online shop, or a guest post syndicated to another site. A canonical tag tells search engines which address to treat as the original. WordPress core already prints a self-referencing canonical on single posts and pages through rel_canonical(). SEO plugins let you override it per post: Rank Math, for example, has a canonical URL field in its Advanced tab that defaults to the post’s own address. Leave it alone unless the post really is a copy of another page.
It is a signal, not an order
Google describes rel="canonical" as a strong signal, alongside redirects, sitemaps and other clues, and says it may choose a different page as canonical than you do. In practice it gets ignored when the two pages are not really duplicates, or when your other signals contradict it: the sitemap lists the other URL, or your internal links all point at the other version.
To see what Google decided, run the URL through Search Console’s URL Inspection tool, which shows the user-declared canonical next to the Google-selected one.
Common misuse
- Using a canonical to “merge” two different articles that compete for the same search. For true keyword cannibalization, a 301 redirect after merging is usually the cleaner fix.
- Pointing every paginated archive page at page 1, which hides the posts on page 2 onward.
- Linking internally to the non-canonical version, which sends mixed signals.
The guide to merging, redirecting or retargeting competing posts sets out when a canonical is the right tool and when it isn’t.