Duplicate, Google chose different canonical than user: why Google overrules your tag, and the fix

Google's Page indexing help puts it in one line: "This page is marked as canonical for a set of pages, but Google thinks another URL makes a better canonical." So Google indexed the URL it chose, not yours. You told Google which URL to index, and it disagreed.

The help then explains what the disagreement means. Google thinks the page you tested isn't a duplicate of the canonical you declared; it thinks it's a duplicate of the one Google selected. And there's a rule that settles a lot of cases on its own: "If the user-declared canonical is not similar to the current page, then Google won't ever choose that URL as canonical." A canonical tag pointing at a different page, like every post pointing at the home page, can't win.

It's the sibling of Duplicate without user-selected canonical, where no canonical was declared at all. Here one was, so the job isn't adding a tag. It's finding out why the tag lost.

Sometimes Google's the one being sensible. Its troubleshooting page asks you to think about whether the Google-selected canonical "makes more sense than your preferred canonical URL for your users coming from Google Search." If the two pages really are the same and Google picked the cleaner or better-linked URL, the fastest fix is to agree: point your canonical at Google's choice, or 301 the weaker URL into it and move its links over.

Check the performance data before you force the issue. In one of my case studies, Search Console reported both the www and non-www homepages: the non-www one at average position 13.2 and the www one at 5.3, while the site canonicalised to non-www. That doesn't mean the declared host was wrong, but it's why consolidating it had to be measured on both URLs through the change.

The other honest answer is that the pages aren't duplicates and shouldn't be. Two location pages grouped together need content that differs clearly, and Google says pages split out of a cluster faster when the difference is clear and significant. A tag can't do that part.

For a handful of important URLs, inspect each one and press Request Indexing. Google's troubleshooting page says to reserve it for your most important URLs because it's subject to quotas, and that even after a content fix Google might hold pages in a duplicate cluster for up to two weeks. More detail on reading those results is in how I use the URL Inspection tool.

For the whole issue, fix every instance and press Validate fix in the Page indexing report. Google says validation typically takes up to about two weeks and sometimes much longer, and that validating against a sitemap of only your most important URLs can finish faster than validating every affected URL. When it passes, inspect the URLs again: the proof is Google-selected canonical matching your declared one, not a shrinking count.