Canonicals and Duplicate Content
The same page on your site probably has several addresses. Not copies you made — the same page, reachable more than one way. Google has to pick one of them to show, and you can either tell it which, or leave it guessing. The word for the one it picks is canonical, and it turns up unexplained in most SEO advice.
What duplicate content actually is
Duplicate content means substantially the same content available at more than one URL. Most of it is accidental and nobody made a copy of anything. A single product page can easily exist at all of these:
http://shop.com/kettleandhttps://shop.com/kettlehttps://www.shop.com/kettleandhttps://shop.com/kettlehttps://shop.com/kettle/with a trailing slashhttps://shop.com/kettle?ref=newsletter— the same page with a tracking parameter on the endhttps://shop.com/kitchen/kettle— the same product filed under a second category
Six addresses, one kettle. A canonical URL is the one you nominate as the real one. The others still work for visitors; they simply stop competing with it in search.
Before going further, one thing worth saying plainly: there is no duplicate content penalty. Google's documentation says duplicate content on a site is not grounds for action against it unless the intent is to deceive. Ordinary duplication is a tidiness problem, not a moral one.
How Google handles it
When Google finds several URLs with the same content, it groups them and picks one to index and show. It uses several signals to choose: redirects, the addresses you list in your sitemap, which version your own internal links point at, and the canonical tag.
That tag lives in the page's <head> and looks like this:
<link rel="canonical" href="https://shop.com/kettle">
Here is the part people get wrong: the canonical tag is a strong hint, not an instruction. Google can and does choose a different canonical when your other signals contradict the tag — for instance if your internal links all point somewhere else, or the tag points at a page that redirects. The tag is one vote, cast alongside everything else your site is saying.
The consequence of leaving it to chance is not a penalty. It is that the signals for one page get split across several addresses, and Google may index the version with the tracking parameter rather than the clean one.
What to do
- Pick one version of your domain —
https, and eitherwwwor not — and permanently redirect the others to it. This alone removes most accidental duplication. - Give every page a canonical tag pointing at itself. A self-referencing canonical is normal and recommended, and it means a page arriving with a parameter attached still names the clean address.
- Point canonicals at a URL that returns a page, not at one that redirects or errors.
- Make your internal links use the canonical address. Contradicting your own tag is the most common reason Google ignores it.
- Check a live page: open it, view source, and search for
rel="canonical". It takes ten seconds and most people have never done it.
Common mistakes
Every page canonicalised to the homepage. A misconfigured theme or plugin does this, and it tells Google that your entire site is one page. It is the single most destructive canonical error there is, and the site looks completely normal while it happens.
Combining a canonical with a block in robots.txt. Google has to fetch the page to read the tag. Block the page and the tag is never seen.
Using a canonical where a redirect belongs. If the old address should genuinely no longer be used, redirect it. A canonical keeps both addresses working, which is right for parameters and category paths and wrong for a page you have actually moved.
Sources: Consolidate duplicate URLs with canonicals and What is URL canonicalization.
Was this useful?
Counts appear once there are 5 votes.
Related posts
Structured Data, Explained Simply
The code behind star ratings and breadcrumbs in search results. What it makes you eligible for, what it never promises, and the one rule that loses it for you.
ReadCore Web Vitals, Explained Simply
"Make your site faster" turned into three specific numbers you can measure, with Google's published thresholds and the usual causes of a bad score.
ReadSitemaps and robots.txt
Two small files that decide what Google may look at and what it knows exists. They do opposite jobs, and one of them can remove your whole site from Google.
ReadTechnical SEO for Founders
A technical SEO guide for founders: the handful of fixes that actually move rankings (indexing, canonicals, speed, structure) and the things you can safely drop.
Read