When the same or very similar content exists at more than one URL, search engines face a choice: which version should they index and rank? Without guidance, they might choose the wrong version — or split ranking signals across multiple versions. The canonical tag is your instruction to Google about which version is the definitive one.
Why do duplicate URLs exist on websites?
Duplicate URLs arise through several common mechanisms:
- URL parameters — tracking parameters, session IDs, and sorting options create multiple URLs showing the same content: /page?utm_source=email and /page?sort=price
- HTTP and HTTPS versions — both http://yoursite.com and https://yoursite.com may be accessible
- www and non-www versions — yoursite.com and www.yoursite.com may both respond with content
- Trailing slashes — /page and /page/ are technically different URLs
- Pagination — /blog and /blog/page/2 might have similar introductory content
- Mobile versions — m.yoursite.com may duplicate desktop content
What is a self-referencing canonical and why does every page need one?
A self-referencing canonical is a canonical tag on a page that points to itself:
<link rel="canonical" href="https://yoursite.com/this-page">Even on pages where no obvious duplicate exists, a self-referencing canonical is best practice. It explicitly tells Google "this exact URL, with this exact protocol, is the definitive version" — preventing Google from choosing between https:// and http://, www and non-www, trailing slash and no trailing slash variants. Without it, Google may choose a different version than the one you intend to rank.
Is a canonical tag a guaranteed instruction to Google?
No — Google treats canonical tags as strong hints, not absolute directives. If your canonical tag contradicts other signals (your sitemap, internal links, or hreflang), Google may choose a different canonical than the one you specified. For canonical tags to be followed reliably, they should be consistent with all other canonicalisation signals on your site.
When do you specifically need a canonical tag beyond self-referencing?
E-commerce faceted navigation — /products?colour=red&size=M should have a canonical pointing to /products/ (the clean category page). Without it, every filter combination generates an indexable URL creating thousands of near-duplicate pages.
Print-friendly versions — /page?print=1 should canonical to the main page URL.
Syndicated content — if you allow other sites to republish your content, ask them to add a canonical tag pointing back to your original URL. This ensures Google credits your site as the original source.
Pagination — each paginated page (/blog/page/2) should have a self-referencing canonical (not canonicalising to page 1, which would indicate page 2 has no unique content worth indexing).
What are the most common canonical tag mistakes?
- Canonicalising to a noindexed page — tells Google to treat a page you want as the canonical, but also tells it not to index it. Contradictory.
- Canonical and redirect pointing to different URLs — if /page-a canonicals to /page-b but /page-a redirects to /page-c, the signals conflict.
- Using relative rather than absolute URLs — canonical tags should always use the full URL including protocol: https://yoursite.com/page not /page
- Canonical chains — Page A canonicals to Page B which canonicals to Page C. Google follows one hop; multi-hop canonicals may not work as intended.
Summary
- Canonical tags tell Google which version of a page to index when duplicates exist — they are hints, not absolute instructions
- Every page should have a self-referencing canonical to prevent protocol and subdomain duplicate issues
- Use additional canonicals for URL parameter variants, print versions, and syndicated content
- Canonical and redirect signals must be consistent — contradictions cause Google to choose its own canonical
- Always use absolute URLs in canonical tags and avoid canonical chains
Continue reading: How do you do SEO for a restaurant or cafe?