Canonical Tags and Duplicate Content: A Plain-Language Guide
A canonical tag tells search engines which address of a page is the preferred one when the same or very similar content can be reached at several addresses. Duplicate content is usually not a penalty, but it splits signals and wastes crawling, so pointing everything to one main version is a good habit for any site.
What duplicate content is
Duplicate content means the same page, or pages that are almost the same, can be opened through more than one address. It is common and often accidental. The page itself is fine, but search engines must choose which address to show, and that choice may not be the one you would pick.
Where duplicates come from
- The address works with and without www.
- Both the HTTP and the HTTPS versions load, as covered in the HTTPS guide.
- An address works with and without a trailing slash.
- Tracking or sorting parameters create new addresses for the same page, as described in the URL structure guide.
- A print version or a mobile version lives at its own address.
- Similar pages exist for different products or places with nearly the same text.
Why it matters
Choosing a preferred address has four practical benefits. You control which address appears in search results. Signals such as links are gathered on one page instead of being split. Tracking is simpler when one page has one address. And crawlers spend their time on pages that matter instead of on copies.
Ways to indicate the preferred address
Search engines rank the methods by how strong a signal each sends, and they work best when combined consistently.
- Redirects. The strongest signal. A permanent redirect sends visitors and crawlers from the duplicate to the main address. See the redirect guide.
- A canonical link in the page. A line in the page header names the preferred address. The duplicate stays reachable, but search engines understand which is the main version.
- The sitemap. Listing only preferred addresses in the sitemap is a simple, supporting signal.
None of them is required. If you do nothing, the search engine picks a canonical address on its own. Specifying one simply gives you control. For files that are not pages, such as PDFs, the preferred address can be named in a header sent by the server instead of in the page.
How to add a canonical tag
If you use a content management system or a website builder, look for an SEO setting for each page, which often fills in the tag automatically. If you edit the code, add one line in the head of the page with the full address of the preferred version. Use the full address including the domain, and choose a single method for a page instead of several.
What not to do
- Do not block duplicates with the robots file as a way of choosing a canonical address. The crawler then cannot see the page to read the signal.
- Do not send conflicting signals, such as a canonical tag pointing to one address and a redirect to another.
- Do not point the canonical tag of a page to an unrelated page.
- Do not use a no-index instruction to choose a canonical address, because it tells search engines to leave the page out entirely.
- Avoid fragments in canonical addresses.
A page can name itself as canonical, which is a good habit. Linking internally to the preferred versions also helps, so menus and buttons should always use the main address.
Check your result
Open Search Console and inspect a page address. The report shows which address the search engine chose as canonical and whether it is the one you declared. If they differ, look for conflicting signals: a redirect, a different link in the sitemap or internal links to another version. The tool is explained in the Search Console guide.
Duplicates on the same site
Sometimes the duplicate is not a technical copy but an editorial one. Two articles that answer the same question compete with each other. The better fix is to merge them into one stronger page and redirect the weaker one. This is part of keeping a blog tidy, as discussed in the guide to refreshing old posts.
A worked example
Imagine a guide that can be opened at four addresses: with and without www, and with a tracking tag added by a newsletter link. To a visitor these are one page. To a search engine they are four candidates. The fix is a short chain of decisions: choose one main address, redirect the www and non-www versions to it, add a canonical link that names the main address on the page itself, and list only that address in the sitemap. The tracking version still opens for the people who click the newsletter, and the canonical link tells search engines to treat it as the same page.
Duplicate content across languages and countries
Pages that have the same text for different countries or languages are a special case. Search engines use separate language annotations to understand that the versions belong together, and they prefer addresses that belong to such a group. Keep these annotations consistent with your canonical tags, so each language version names itself as canonical and does not point to another language.
A short audit you can do in an hour
- List the ways your homepage can be opened and make sure they all end up at one address.
- Pick five important pages and inspect each in Search Console to see the canonical address chosen.
- Search your site name together with a distinctive phrase from a page and check whether more than one address appears.
- Review the sitemap and remove addresses that redirect or that you do not want shown.
FAQ
Is duplicate content a penalty?
Normally no. Search engines group similar pages and pick one. Problems come from wasted crawling and from the wrong address being shown.
Do I need a canonical tag on every page?
It is not required, but a self-referencing tag on each page is a clean habit and many platforms add it automatically.
What is a self-referencing canonical?
It is a tag on a page that names that page's own address as the preferred one.
How do canonical tags differ from no-index?
A canonical tag suggests which version to prefer. No-index asks search engines not to show the page at all.
Can I use a canonical tag across different domains?
Yes, for content that is syndicated or republished elsewhere, though search engines treat it as a hint and may choose differently.
Your next step: type your homepage into the browser with and without www, and with HTTP and HTTPS. All four should end up at one address. If they do not, set up redirects.