A canonical tag (rel="canonical") is an HTML element placed in a page's <head> that tells search engines which URL is the "master" version of that content. It prevents duplicate-content issues when the same or similar content is accessible at multiple URLs.
When You Need It
Product pages accessible via multiple category paths. HTTP and HTTPS versions of the same page. Pages with URL parameters (tracking codes, sort orders, filters). Syndicated content that appears on multiple domains. AMP and non-AMP versions of articles. Any situation where Google might see two pages as duplicates and split ranking signals between them.
How It Works
The canonical tag points from the duplicate to the original. Search engines consolidate ranking signals — backlinks, engagement, freshness — to the canonical URL. The duplicate may still be crawled but won't compete with the original in search results.
Common Mistakes
Self-referencing canonicals are best practice (every page should canonical to itself). Canonical tags pointing to non-existent pages. Canonical chains (A → B → C). Canonicalizing pages with substantially different content. Conflicting canonical and redirect signals.
AI Implications
AI engines also encounter duplicate content across crawled pages. Proper canonical signals help ensure that AI training pipelines ingest the intended version of content rather than a parameter-cluttered duplicate.