What is canonical url?
A canonical URL is the address you nominate as the single official version of a page, so search engines know which one to index when the same content appears at more than one address.
Worked example, not a result
One product reachable at several addresses
Say a crawl of a store finds 9,800 URLs for a catalogue of only 1,400 products. The extra addresses are not new content. They are the same products reached through different routes: a product inside each collection it belongs to, the same product with a tracking parameter on the end, the same product with a sort order applied, and a variant with a colour parameter.
A search engine seeing seven addresses for one product has to decide which is the real one. Left to guess, it may pick a version buried inside a collection, or split the signals from links and traffic across several addresses so none of them is as strong as the single page should have been.
A canonical tag settles it. Each of those addresses names the clean product URL as the official one, and the value they have gathered consolidates into that page.
The counts above are an example.
Why it matters
Duplicate addresses are one of the most common reasons an ecommerce site underperforms in search, and stores rarely notice because every version looks fine to a human. It is a visitors lever problem hiding in the site structure rather than in the content.
Ecommerce platforms generate these duplicates by design. Collections, filters, sort orders, variants, tracking parameters and print views all produce alternative addresses for the same content, and a catalogue of a few thousand products can easily present tens of thousands of URLs.
For a store owner the useful check is a crawl. Compare the number of URLs found against the number of products and collections that actually exist. If the first is several times the second, the site is diluting its own pages, and canonical tags plus sensible internal linking usually recover more search traffic than any amount of new content.
Common mistakes
- Pointing every canonical at the homepage. That tells search engines the product pages are not worth indexing, and they will act on it.
- Canonicalising pages that are genuinely different. Two distinct products or two distinct collections each need their own page, and merging them removes one from the results.
- Setting a canonical to a page that redirects or no longer exists. The signal is ignored and the duplicates stay in the index.
- Assuming a canonical tag is an instruction. It is a strong hint, so internal links, sitemaps and navigation should point at the same version to support it.
Related terms and answers
Your plan is free.
A person reads it. You keep the plan. No logins, no call first.