Refract Group2026
Glossary
Canonical URL
A canonical URL is the one address a site asks search engines to index when the same page answers at several. It is declared with a canonical link in the page’s head, and Google treats it as a hint it can overrule.
refract.group/About?utm_source=x
<head>
<link rel="canonical"
href="https://refract.group/about">
<meta property="og:url"
content="https://refract.group/about">
</head>
Canonical URL in detail
- Also called
- Canonical tag, rel=canonical, Preferred URL
- Updated
What it is
The same page often answers at more than one address. Capital letters in a shared link, a trailing slash, a tracking parameter such as ?utm_source=, the http and https versions: to a search engine each is a separate URL that happens to return the same content. Canonicalisation is how the search engine picks one of them to index and show, and gathers the signals from the rest onto it.
You state your choice with a link element in the head, rel="canonical", holding the absolute URL you prefer. Most pages name themselves. Redirects and the sitemap send the same message in other ways, and Google weighs them all together. It calls the canonical you declare a hint, and may pick a different page when the signals disagree.
Why it matters
Duplicates split what a page has earned. Links to one variant, clicks on another and shares of a third are counted against different URLs, and the version Google indexes may be one with a campaign tag in it, or one you never meant to publish. A canonical on every page, with internal links and a sitemap that use the same form, settles which address is the page. It is cheapest to set when a site is built, in the one layout every page shares.
How we do it
We set one rule and apply it to every page from a single place. The canonical is the production address plus the path, lowercased, with no trailing slash and no query string. Before that, /ABOUT, /about/ and /about?utm_source=x all answered with the same page and no canonical, so each was a separate page to a search engine.
The Open Graph URL uses the same string, so a share of any variant counts toward one address. Addresses with a trailing slash now redirect permanently to the bare form, and the sitemap lists every page in the canonical form and nothing else. An error page declares no canonical at all.
An example
Say a pricing page at /pricing goes out in a newsletter as /Pricing?utm_source=newsletter. Without a canonical, a crawler that follows the link finds a copy of /pricing at a new address and has to guess which to keep. With one, both versions carry a link naming https://example.com/pricing, the sitemap lists that address and the site’s own links use it. The campaign tag still reaches the analytics, and the search engine indexes one page.
Common mistakes
- Pointing every page at the homepage. A canonical claims two pages carry the same content, so one that names an unrelated page sends a signal the page itself contradicts.
- Writing relative paths. Google asks for absolute URLs in the canonical link, with the protocol and the domain written out.
- Letting the signals disagree. A canonical that names one URL while the sitemap and the internal links use another leaves Google to choose, and it may choose the one you did not want.
- Blocking duplicates in robots.txt. Google says not to use robots.txt for canonicalisation, since a page the crawler cannot fetch cannot show it which URL it prefers.
Canonical URL vs 301 redirect
A canonical URL leaves every version of a page reachable and asks search engines to index one of them, while a 301 redirect sends people and crawlers from the old address to the new one.
| Canonical URL | 301 redirect | |
|---|---|---|
| Visitor sees | The address they asked for | The new address |
| Where it lives | A link in the page’s head | The HTTP response, before any page |
| Use when | Both versions have to stay live | The old address is retired |
| Typical case | Tracking parameters, sorted lists | A page moved in a rebuild |
Questions
What people ask about canonical URL.
No. Google treats a declared canonical URL as a hint. It weighs it alongside redirects, the sitemap, internal links and whether the page is served over https, and may choose a different URL when those disagree or the pages differ. Making every one of those signals name the same address gives yours the best chance.
Yes. With a self-referencing canonical, a page names its own clean URL, so every variant of it, with tracking parameters or different capitals, points back to one address. Google’s site move guidance asks for self-referencing canonicals on new URLs. Added once to the layout every page shares, it covers the whole site.
Yes. A sitemap lists the URLs you want crawled, and Google counts a URL’s presence there as a weak signal toward making it canonical, while the canonical link on the page is a strong one. A sitemap that lists an address whose page names a different canonical sends both signals against each other, so build both from one rule.
Where this shows up in our work
Related terms
Working on this now?
Tell us what you’re building and where it’s stuck. We reply within 12 hours.
Most teams start with a call, then a sprint or an audit.
The people you’d work with, from the first call to launch