Canonical URLs: handling duplicate content cleanly
A canonical URL is the version of a page you want search engines to treat as the definitive one when several URLs show the same or very similar content. The canonical tag is how you point engines to it, keeping duplicate versions from splitting your ranking signals.
Last updated
The duplicate content problem
Sites often end up with the same content reachable at several URLs — with and without a trailing slash, with tracking parameters, printable versions, or the same product under different category paths. To an engine these can look like separate pages competing with each other.
The harm is dilution. Links and relevance that should concentrate on one page get scattered across its duplicates, and the engine may pick the wrong version to show. Consolidating the signals onto a single chosen URL fixes this.
How the canonical tag works
The canonical tag names the preferred URL for a piece of content. Placed on the duplicate versions, and on the preferred page pointing to itself, it tells engines to credit the ranking signals to the one canonical version rather than dividing them.
It is a strong hint rather than an absolute command — engines usually honour it but can override an obviously wrong choice. Used consistently and correctly, it is the standard, low-risk way to manage duplication without deleting useful pages.
Getting it right
The common errors are pointing canonicals at the wrong page, canonicalising every page to the home page, or mixing signals so the canonical contradicts internal links and the sitemap. Each confuses engines about which version truly matters.
The reliable pattern is to choose one canonical URL per piece of content, point all duplicates at it, have it reference itself, and make sure links and sitemap agree. Consistency across every signal is what makes canonicalisation work.
Where Kirti fits
Kirti is being built to detect duplicate and near-duplicate pages during its technical audit and to check that canonicals are present, self-consistent, and pointing at the right versions — proposing corrections for the team to approve.
The agent is in active development and not yet generally available; canonicalisation works the same across sites, so this guidance applies whatever platform you use.