Canonical URL

A canonical URL is a page's declaration of its own real address. It sits in the page's code as a single line naming the one version that should count. It exists because the same page can almost always be reached by several different addresses, and a machine has no way of knowing they are the same page rather than several similar ones.

The duplicates arrive on their own. A site answers on both http and https, with and without the www, with and without a trailing slash. A shop adds a sort or filter to the end of a category address. A newsletter link arrives carrying a tracking tail. Each of these is, as far as a machine is concerned, a separate page that happens to contain identical words. Nobody creates them deliberately, and almost every site has them.

What follows is a slow and invisible dilution. Links and mentions attach to whichever version happened to be shared, so no single address accumulates the standing that all of them together earned. A machine deciding which one to store and cite picks for itself, and it may well pick the one with the tracking parameter still attached. Your own pages end up competing with each other over the same question, and none of them wins cleanly.

The correction is small and largely one-time. Every page names itself with a full, absolute address, using the same host, the same protocol and the same trailing-slash habit that your links and your sitemap already use. A page pointing at itself is the normal case, not a special one. Where a duplicate is deliberate — a print version, a filtered listing — it points at the original instead.

The plain version: this is a business agreeing with itself about its own address. You would not print four variations of your street across four batches of letterhead and expect the post to arrive reliably. The fix is a line of code, but the discipline underneath it is consistency — saying the same thing about yourself everywhere, which turns out to be most of this work in the end.