canonical urls: the one line that stops google splitting your ranking

web · 2026-10-13 · 6 min

google sees your site at four addresses: http and https, with and without www. without a canonical tag it must guess which one counts — and guesses split your ranking across duplicates. one line of HTML fixes it. most small sites lack it.

what the tag says

a canonical tag names the one true URL for a page. every variant (http, www, trailing slash, tracking parameters, print versions) points to it, and search engines consolidate credit there instead of dividing it among copies. it is not a redirect — visitors still land on variants — it is an instruction about which address owns the ranking.

the four-address problem

type your domain four ways and watch: many sites serve the same page on all four with no canonical. google then picks a favorite, possibly a different one per page, and your backlinks point at three addresses while credit lands on one. the fix takes minutes: pick https + one www choice, redirect the rest, canonical everything to itself. the redirect half of this: http vs https, and the not secure warning.

shop duplicates and parameters

filters and sorts multiply one product into dozens of URLs (?color=, ?sort=, ?page=). canonicalize them all to the clean product URL so the product ranks instead of its 30 shadows. this is the single highest-value SEO fix on most small shops, above any keyword work: keywords matter only after consolidation, in seo for beginners, five moves.

FAQ

is canonical the same as redirect?

no. a redirect moves visitors. a canonical tells search engines which URL owns the credit while visitors stay put. use redirects for dead and duplicate addresses; use canonicals for live pages with necessary variants.

can I point canonicals at another page?

only for true near-duplicates. pointing a canonical at an unrelated page reads as a soft deletion of the first — its ranking drains away.