Canonical URL

Shopify stores hit this constantly. A single product can be reachable through its product URL, through one or more collection paths, and through variant or sort parameters appended to the query string, so the same content lives at many addresses. Left alone, that scatters links, crawl budget, and ranking authority across near-identical pages that compete with each other. Search engines then have to guess which version deserves to rank, and they do not always guess the way you would.

The canonical tag points every duplicate at the address you want indexed, and search engines fold the signals from the others onto it. Shopify sets sensible canonicals by default, pointing variant and collection-scoped product URLs back to the clean /products/handle path, which is usually what you want. The tag lives in the head of the duplicate pages and names the master, so the work happens quietly without changing what a shopper sees.

Consider a store selling a linen apron. The same product is reachable at /products/linen-apron, at /collections/aprons/products/linen-apron from the collection grid, and at /products/linen-apron?variant=44821 when a shopper picks the natural colourway. A marketing campaign then adds /products/linen-apron?utm_source=newsletter to the mix. That is four addresses for one product. With each variant and parameter URL carrying a canonical back to /products/linen-apron, reviews, backlinks, and engagement gathered anywhere in that set accrue to the one page you actually want ranking, rather than diluting across four.

There is an answer-engine angle worth being honest about. Tools like ChatGPT, Perplexity, and Google AI Overviews lean on the indexed, canonical version of a page when they retrieve and cite sources. If your authority is split across duplicates, none of them looks like the definitive answer, and a model is less likely to surface or quote you. Consolidating onto one clean URL gives these systems a single, well-signalled page to trust and reference, which matters as more product research starts inside a chat interface rather than a results list.

A canonical is a strong hint, not a command. Google can ignore it if your internal links, sitemap, and on-page signals point somewhere else, so keep them consistent. It also does not replace a redirect. If a URL should genuinely no longer exist, use a 301; reserve rel=canonical for pages that need to stay live but should not be indexed separately. Audit yours by viewing source on a few product and collection URLs and confirming each points where you expect, and check the URL Inspection report in Google Search Console to see which version Google has actually chosen as canonical.