Canonical Tag
Tells Google which version of a page is the “real” one when duplicates exist.
What it means
A canonical tag is an HTML element in a page’s <head> — rel="canonical" — that names the URL search engines should treat as the authoritative version of that content. It exists for one situation: two or more URLs serve the same thing, and you want the ranking signals consolidated onto one of them.
How it affects your SEO
Without a canonical, Google picks its own preferred URL and may index several copies, splitting links, content signals and rankings across all of them. A correct canonical consolidates that authority onto one URL, so the page competes once instead of several times weakly.
What a canonical tag looks like
It is one line in the <head> of the page you want de-duplicated, pointing at the URL you want indexed — never the other way round. Both URLs keep working for visitors; only the ranking signals are merged.
<link rel="canonical" href="https://yourdomain.com/services/roofing" />What a correct one has to satisfy
- The href is an absolute URL — full https:// address, not a relative path.
- The target returns a 200. Pointing at a redirect or a 404 throws the signal away.
- Exactly one canonical tag per page. If two are present, the first wins, and it may not be the one you meant.
- The target is indexable — a canonical pointing at a noindexed URL cancels both directives.
- The target is not blocked in robots.txt; a crawler that cannot fetch the target cannot follow the hint.
Self-referencing canonicals
A page whose canonical points at itself looks pointless and is not. It is the cheapest way to keep tracking parameters, print views, session IDs and sorted filter URLs from being indexed as separate pages, because every variant points back at the clean address.
Add it as a template rule rather than page by page: every indexable page emits its own clean URL as canonical, automatically. That is the difference between a one-off fix and a system that holds when someone shares a link with five UTM parameters attached.
Canonical, redirect, or noindex?
- Same content, several URLs — parameters, sorting, print views, www and non-www: canonical.
- The page has permanently moved somewhere else: a 301 redirect, not a canonical. A canonical does not move visitors.
- The page should never appear in search at all — staging, internal search results, thin utility pages: noindex. A canonical does not remove anything from the index.
- Two genuinely different pages competing for one query: neither. That is a content problem, and canonicalising one of them just hides it.
Why Google sometimes ignores yours
- Your canonical conflicts with other signals — internal links, the sitemap, the HTTP header or the redirect path all pointing somewhere else.
- The tag is injected by JavaScript after load, and the crawler read the raw HTML first.
- The canonical target is a different domain you do not control. Treat cross-domain canonicals as a request, not a guarantee.
- Every paginated page is canonicalised to page one, so pages two and three — and the products on them — never get indexed.
- Language or region variants are canonicalised to each other. Use hreflang instead; canonicals collapse alternates rather than declaring them.
How to check yours
- View source and search for rel="canonical" — confirm it names the clean URL, not a staging host or an http:// version.
- In Search Console, URL Inspection reports "Duplicate, Google chose different canonical" and names the URL Google actually picked. That is the only opinion that counts.
- After a crawl, check every indexable page returns exactly one canonical tag — and that the target of each one is itself indexable.
Canonical Tag — frequently asked questions
Does a canonical tag pass all link equity to the chosen URL?
Yes — a correct canonical consolidates links and content signals onto the preferred URL, so the page ranks as strongly as possible instead of splitting authority across duplicates.
Will Google always honor my canonical tag?
Not always. Google treats canonical as a strong hint, not a command. If the tag conflicts with signals like internal links, the sitemap or redirects, Google may choose a different URL entirely.
Where exactly does the canonical tag go?
In the <head> of the page being de-duplicated, naming the URL you want treated as authoritative. Not in the body, and not on the preferred page pointing at its own copies.
Can a canonical tag hurt my rankings?
Only if it points somewhere wrong. Pointing a good page at a redirect, a 404 or a weaker URL tells Google to ignore the page you wanted ranked — which usually looks like a sudden ranking drop with no crawl error to explain it.
Do canonical tags work across different domains?
They are a hint, and Google sometimes honours them for syndicated content when both sites are verified in Search Console. Treat it as a request: never rely on a cross-domain canonical to solve a duplicate-content problem.
What is the difference between a canonical tag and hreflang?
A canonical consolidates duplicates into one URL. Hreflang keeps genuinely different language and region versions indexed and serves the right one to each audience. Canonicalising language versions to each other removes them from search.
Related SEO terms
Google’s three speed-and-stability metrics: LCP, INP, and CLS.
Structured data that labels content (reviews, FAQ, products) for search engines.
How many pages Google will crawl on your site per day.
A file that tells crawlers what they may and may not crawl.
Is this hurting your rankings?
Run a free 60-second audit and find out where canonical tag — and dozens of other factors — stand on your site.
Get my free SEO score