Canonical URL: How to Tell Google Which Page Is the Main One

Where duplicate pages come from

Duplicates are rarely created on purpose. They appear on their own because of CMS settings, catalog filters, ad tracking tags and different ways of reaching the same page. Site owners usually notice only when traffic starts splitting oddly between similar addresses.

The most common sources:

• URLs with sorting and tracking parameters, such as UTM tags from Telegram and Instagram campaigns;
• one product reachable under different addresses in several categories;
• filter pages, where combinations of size, colour and price produce dozens of near-identical versions;
• www and non-www, http and https, trailing slash and no slash;
• your articles republished on other sites or platforms.

How a canonical URL works

It is set with a rel=canonical tag in the head of the page. The tag holds the full link to the version that should count, including protocol and domain, for example https://example.uz/catalog/sneakers/. Every similar page points to that address, so the search engine knows which version is the main one and treats the rest as copies.

A good habit is to add a canonical tag to every page, including the main one, pointing at itself. Even if someone later appends a campaign parameter to the address, the search engine still gets a clear reference.

A hint, not a command

It is important to understand that canonical is a recommendation. Search engines take it into account but may pick a different address when signals conflict. For example, internal links across the site lead to one version while the tag points to another. Or the pages differ so much that they no longer look like duplicates.

That is why the tag works best when everything else says the same thing: the menu, links inside articles, the sitemap and redirects all lead to the same main address.

Canonical or a 301 redirect

If users do not need the duplicate page, a permanent 301 redirect is more reliable: both the visitor and the crawler land on the right address straight away. Typical cases are moving from http to https or merging www and non-www versions.

Canonical fits when both versions must stay accessible. A shopper sorts the catalog by price and needs that page, but it should not compete in search with the main category.

Common canonical mistakes

Most problems come from a tag that was added automatically and never checked again. What we see most often:

• the canonical points to a page blocked from indexing or returning a redirect;
• chains, where page A points to B and B points to C;
• every page in a section points its canonical to the homepage;
• the wrong protocol or a foreign domain in the address;
• all paginated pages pointing to page one even though their content differs;
• several canonical tags with different addresses on a single page.

How to check your setup

Run a crawler: it lists the canonical of every page. Make sure the address in the tag opens with a 200 status, can be indexed and does not redirect elsewhere. Then open URL Inspection in Search Console, which shows which address Google chose as canonical. If it differs from yours, look for conflicting signals, most often internal links.

Where to start

Export the list of pages with parameters and filters, decide on a main address for each group and check that internal links and the sitemap lead to it. On a large site, such as a catalog with thousands of products, begin with an audit: canonical errors at that scale are hard to spot by hand.

If you want help, markdev offers SEO optimisation in Uzbekistan, including duplicate analysis and canonical setup. Tell us about your site and we will suggest where to begin.