Canonical tag mistakes that point search at the wrong URL

Check where each template points

Every indexable page should canonicalise to itself, including the protocol, host, and trailing-slash style you have chosen. A sitewide canonical to the homepage collapses the site into one URL.

Also look for canonicals that point at redirected URLs, at URLs blocked in robots.txt, or at a URL with a different intent from the page the user is reading.

• Use absolute URLs.
• Match the canonical host to the preferred host.
• Do not combine a noindex page with a canonical to a different URL and hope for a tidy outcome.

Use canonicals for true variants only

Tracking parameters, print views, and http/https duplicates are fair uses. Separate articles that happen to share a topic are not. If the body answers a different question, it needs its own canonical.

When two pages both claim to be the canonical of each other, or a page canonicalises to a URL that canonicalises elsewhere, flatten the chain.

Confirm the preferred URL is the one that gets links

Internal links, the sitemap, and the canonical should name the same URL. If navigation links to one version and the tag prefers another, you are splitting the hint.

After a fix, inspect the URL in Search Console and check which URL Google selected. The declared canonical and the chosen canonical should agree.

Can a canonical tag move rankings from one page to another?

It can consolidate duplicates when the pages are genuinely the same document. It is a poor way to transfer rankings between pages that answer different questions.

Should paginated pages canonical to page one?

No. Each page in a real sequence is a different document. Canonicalising them all to page one hides the later items. Give each page a self-referencing canonical.