Using Canonical Tags Correctly
Enter a URL and this tool produces the rel=canonical link tag to paste into your page's head. Simple output, but the concept behind it causes a surprising amount of confusion and a fair amount of accidental damage.
A canonical tag tells search engines which URL you consider the authoritative version of a page when the same or very similar content is reachable through more than one address. It is the standard answer to duplicate content that arises from URL variations rather than from copied writing.
Where duplicate URLs come from
Most duplicate content is accidental and structural rather than anything anyone wrote twice. The same page routinely becomes reachable at several addresses through entirely normal site behaviour.
- Protocol and host variants — http and https, www and non-www, all serving identical content.
- Trailing slashes — /page and /page/ treated as distinct URLs by search engines.
- Tracking parameters — ?utm_source=newsletter creating a new URL for every campaign.
- Sort and filter parameters — ?sort=price generating many URLs for one product list.
- Session IDs and print versions still produced by some older systems.
- Syndicated content republished on another domain with permission.
What the canonical tag does
It consolidates. When several URLs carry a canonical pointing at one address, search engines treat that address as the version to index and attribute the ranking signals from all the variants to it. Without it, those signals get split across the duplicates and none of them ranks as well as one consolidated page would.
It is a hint, not a command. Google usually respects a clear canonical, but it makes its own judgement and can choose a different URL if your signals contradict each other — for instance if your canonical points one way while your internal links, sitemap and redirects all point another.
Consistency is therefore what makes canonicals work. The canonical URL should be the same URL you link to internally, the same one in your sitemap, and the one your redirects resolve to.
Self-referencing canonicals
Every indexable page should carry a canonical pointing at itself. This sounds redundant and is genuinely useful: it removes ambiguity when a page is reached with tracking parameters appended, which happens constantly with email and ad traffic.
Without a self-referencing canonical, ?utm_source=newsletter can be indexed as a separate URL from the clean version. With one, the parameter version explicitly declares the clean URL as authoritative and the signals consolidate.
Use the absolute URL including protocol and domain. Relative canonicals are technically permitted but resolve unpredictably and are a common source of subtle errors.
Canonical versus redirect versus noindex
These three solve different problems and are frequently used interchangeably by mistake.
- Canonical — the duplicate URL should stay accessible to users, but only one version should be indexed. Filters, tracking parameters, syndicated copies.
- 301 redirect — the old URL should no longer be reachable at all. Moved or merged pages. Trace existing ones with the Redirect Chain Checker.
- noindex — the page should be reachable but never appear in search. Thank-you pages, internal search results, thin utility pages.
Mistakes that cause real damage
The worst is canonicalising every page to the homepage. It happens when someone hardcodes a canonical into a shared template, and the effect is telling Google that no page on the site other than the homepage deserves indexing. Traffic collapses and the cause is invisible unless you view source.
Canonical chains are the next problem: page A canonicals to B, which canonicals to C. Google may follow the chain or may ignore it entirely. Point every duplicate directly at the final canonical URL.
Combining a canonical with a noindex on the same page sends contradictory instructions — one says consolidate signals here, the other says keep this out of the index. Pick one. Similarly, canonicalising to a URL that redirects, is blocked in robots.txt, or returns a 404 makes the tag unusable.
Finally, canonicals should point at genuinely equivalent content. Pointing a product variant page at an unrelated category page is not consolidation, and Google will simply disregard it.
Checking what a page currently declares
The On-Page SEO Analyzer reports the canonical tag found on any live URL, which is the fastest way to catch a template canonicalising everything to one place. Search Console's Page Indexing report is the other place these problems surface, usually as “Alternative page with proper canonical tag” or “Duplicate, Google chose different canonical than user”.
That second message is worth taking seriously. It means Google looked at your canonical, disagreed, and picked something else — usually because your other signals point elsewhere. The fix is aligning internal links and sitemap entries with the canonical you actually want.
Frequently asked questions
Should every page have a canonical tag?
Every indexable page should have a self-referencing canonical. It costs nothing and prevents tracking-parameter versions of your URLs being indexed separately from the clean version.
Is a canonical tag the same as a 301 redirect?
No. A redirect makes the old URL inaccessible and sends everyone to the new one. A canonical leaves both URLs working for users while telling search engines which to index. Use redirects for moved pages, canonicals for duplicates that must remain reachable.
Can I canonicalise to a page on another domain?
Yes, cross-domain canonicals are supported and are the standard approach for syndicated content — the republished copy points back to your original.
Why is Google ignoring my canonical tag?
Canonicals are hints, not directives. Google overrides them when other signals disagree — internal links, sitemap entries or redirects pointing at a different URL, or a canonical target that redirects, 404s, or holds substantially different content.
Should the canonical URL be absolute or relative?
Always absolute, including protocol and domain. Relative canonicals resolve unpredictably depending on the page's location and are a frequent source of hard-to-spot errors.