Why might one page have several addresses?
A product page may be reachable through its ordinary address and a version with campaign parameters. A website can also accidentally publish the same content under different hostnames or paths.
Imagine a stationery shop sharing links to the same notebook in several campaigns. The tracking details may differ even though visitors see the same product. The shop needs a consistent preferred address for that product, rather than treating every tracking variation as a separate article.
What does the canonical tag do?
A rel="canonical" link in the page head points to the preferred URL. Google treats it as a strong signal when choosing a representative page, but it can select another URL. Read Google’s canonicalisation guidance.
The tag does not move a visitor. Both addresses may remain accessible in a browser. If the old page should send people to a replacement, consider a redirect instead.
Choose the version you actually want to maintain
Start by comparing the pages. Do they contain essentially the same information, or do they answer different questions? A notebook product page and a guide to choosing paper are related, but they are not duplicates.
For the preferred page, check that it works, contains the intended content and is suitable for public search. Use that same address in internal links and the sitemap. Avoid giving conflicting preferences in different places.
A self-referencing canonical identifies the preferred page itself. It should use the correct public address rather than an editor preview or local development URL.
Common mistakes to avoid
- Pointing every page at the homepage.
- Choosing a destination that returns an error.
- Using a canonical to combine pages with substantially different purposes.
- Leaving an old hostname in the template after a move.
- Confusing a preference signal with a rule to exclude a page from search.
Be especially careful with shared templates. One incorrect setting can appear on many pages even when each looks normal to a visitor.
How to investigate a concern
Record the URLs involved and why you believe they are duplicates. Inspect their declared canonical values, then use Search Console’s URL Inspection information to compare the declared and Google-selected versions where available.
Correct the underlying inconsistency before requesting another review. Allow time for recrawling and processing; changing a tag does not instantly replace a search result.
For the broader process, read crawling vs indexing and our sitemap guide.