Definition
What is a canonical tag?
A canonical tag is a line placed in a page's code (the rel=canonical attribute) that tells Google which URL to favour among several similar versions of the same content. Google treats it as a suggestion: it chooses the representative URL itself, also taking redirects, the sitemap and internal links into account. It consolidates signals on one URL without removing the others from the site.
How Google treats it
Google describes canonical URL selection as a deduplication process: among a set of duplicate pages, it selects the most representative one and shows only that one in its results. The canonical tag is one of the ways to express your preference. None is mandatory: without a signal, Google chooses on its own the version it judges most relevant.
Your tag therefore remains a suggestion. Google specifies that the methods reinforce each other when they agree: a tag backed by internal links and a sitemap pointing to the same URL carries more weight than a tag alone. The pages not selected remain accessible to visitors. Google may keep crawling them, less often over time.
When to use it and when to choose something else
The canonical tag fits when several URLs must stay accessible while showing the same content: a page with or without ad tracking parameters, a list sorted differently, a product variant that stays an option (see the distinct page threshold). Elsewhere, another tool does the job better:
- A 301 redirect, when the old URL no longer has a reason to exist (merged page, changed URL). The visitor and the crawler land directly on the right page.
- A noindex tag, when a page must stay online without ever appearing in results.
- Blocking in the robots.txt file, when URLs should not even be crawled, such as a catalogue's filter combinations. For that case, Google describes the canonical as generally less effective in the long term. A blocked page is not read: neither is its canonical tag.
The mistakes that cancel it
- Two tags on the same page. Google ignores all of them. It happens when a plugin and a template each add one.
- Pages that are too different. Google recommends that a large part of the duplicate page's content be present in the canonical version. Otherwise, the tag may be ignored.
- A wrong URL. An automatic generation error pointing to the wrong page can go unnoticed for months.
- Pagination. Pointing page 2 of a list to page 1 is not a correct use according to Google: they are not duplicate pages. The content of the following pages may then be ignored.
- Conflicting signals. A tag naming one URL while internal linking or the sitemap name another weakens the signal.
- A canonical from one language to the other. A French page and its English translation are not duplicates. Each names itself as canonical, and the two declare each other with the hreflang tag.
Checking what Google kept
Your signal and Google's decision are two different things. In Search Console, the URL Inspection tool shows, for a given URL, the canonical you declared and the one Google selected. The Page indexing report also sets apart duplicate pages for which Google chose a different canonical than yours: that is the status to watch.
Compare the two values on about twenty representative pages. A gap that recurs across a whole family of pages points to a template problem rather than a page problem. After a fix, allow one to three weeks for a canonical change to be taken into account, according to the timings Google presented in October 2026.
Fictional example: a store's Winter boots category answers at three URLs: /winter-boots/, /winter-boots/?utm_source=newsletter and /winter-boots/?sort=price. The last two carry a canonical tag pointing to /winter-boots/. The main page carries the same tag pointing to itself. The menu and the sitemap also point to it.
Google receives three signals that agree: it consolidates links on a single URL and shows only that one. If the sorting plugin added its own tag, the page would carry two, and Google would ignore both.
We read the canonical tag as a useful fix, never as an architecture decision. Before adding one, we look for why several URLs show the same content. On a catalogue, the real decision is choosing which variants and which filters deserve a page: a tag placed on a poorly thought-out architecture moves the problem without solving it.
After the fix, we check in Search Console which canonical Google kept, rather than assuming it was followed. On automatically generated catalogues, two mistakes can be found in an hour: duplicate tags and conflicting signals.
Not to be confused with
- Distinct page threshold
- The distinct page threshold decides in advance whether a product variant deserves its own URL. The canonical tag comes in after the fact, when similar versions already exist. Nothing guarantees it will be followed.
- 301 redirect
- A 301 redirect sends the visitor and the crawler to another URL: the old page is no longer visible. A canonical tag leaves both pages accessible and only suggests which one to show in results.
- Hreflang tag
- The hreflang tag links the versions of a page in different languages or countries, which are not duplicates. The canonical tag picks one URL among duplicate pages in the same language.
Related concepts
- Distinct page threshold
- Faceted navigation
- Indexing
- Crawled, currently not indexed
- Crawl budget
- Path-parameter boundary
- Language cannibalization
- Technical SEO
Further reading
- Product variants and SEO: when should you create a separate page?
- Faceted navigation: managing filters to protect your crawl budget
- Technical SEO: finding crawl and indexing problems
- Local SEO: creating distinct pages for each business location
- How long does Google take to index a site? The timings Google gave in 2026
- eCommerce SEO: how to structure URLs for a large catalogue
Related services
Frequently asked questions
Can Google ignore my canonical tag?
Yes. Google states that no method is mandatory and that it chooses the representative URL itself. It can also ignore a tag when the pages are not similar enough, or when internal linking and the sitemap name another URL. The URL Inspection tool in Search Console shows which canonical it actually kept.
Does every page need a canonical tag?
It is not mandatory: without a signal, Google picks a representative URL on its own. A tag that names the page itself remains a simple precaution on a site where parameters (ad tracking, sorting, filters) create copies of the same URL. It only helps if it agrees with your internal links and your sitemap.
Canonical tag or 301 redirect: which should I choose?
The redirect, when the old URL no longer has a reason to exist: merged page, changed URL. The canonical tag, when both URLs must stay accessible to visitors, such as a sorted list or a page with tracking parameters. A redirect is an instruction the visitor follows too; the canonical remains a suggestion made to Google.
