Canonical URL
The one address you declare as the real home of a piece of content, when the same content is reachable at several. Without it, machines have to guess which version to trust and quote.
Updated August 21, 2026
Duplicate addresses appear on their own: with and without www, with tracking parameters, with a trailing slash, in a print view. Each one splits the evidence for that page. A canonical tag consolidates them into a single address, which is exactly what a model needs when it decides which URL to cite. It is a small tag with a large effect on whether your strongest page reads as one strong page or four weak ones.
Related terms
- Thin content and doorway pages. Mass-produced pages with little unique value, for example the same page repeated with only a city name swapped. Search engines demote these, and AI systems ignore them. The fix is genuine, unique content per page, not cosmetic variation.
- Crawler (bot). An automated program that fetches web pages. Search engines and AI systems use crawlers to read the web. Crawlers read the raw page and often do not run all of a site's scripts.
- Retrieval. The step where a system finds and fetches your page. If a page is blocked, unreachable, or clearly not about the topic asked, it is never retrieved, and nothing else matters.