Canonical tag

It exists because the same page is often reachable at several addresses — with and without a trailing slash, with tracking parameters, with and without www, in a paginated view, in a print version. Left alone, search engines have to pick one, and they may not pick yours.

It is a hint, not a command. Google can and does ignore a canonical it disbelieves — most often when the two pages are not actually duplicates, or when your internal links and sitemap point at a different URL than the tag does. Consistency across those three is what makes it stick.

The mistakes are expensive and quiet. A site-wide canonical pointing every page at the homepage removes the whole site from results. A canonical on a paginated series pointing at page one loses everything on the later pages. A self-referencing canonical on every page is the safe default, and it is worth checking after every template change.