Duplicate, Google chose different canonical than user

Read the status as a disagreement, not a broken tag

‘Duplicate, Google chose different canonical than user’ means Google agrees the URL is a duplicate and disagrees about which copy should represent it. The page declared one canonical. Search selected another. The declared URL can leave the index while the chosen URL keeps the links, the impressions, and the right to be quoted.

This is a different SEO issue from a canonical tag that points at the homepage by mistake, and a different one from ‘Crawled — currently not indexed’, where Google fetched a URL and left it out. Here the pages were judged to be the same document, and your preference lost. ‘Alternate page with proper canonical tag’ is the healthy twin: you asked for consolidation and Google agreed. This status is the one teams search for when the tag and the indexed URL have split. It is also a GEO issue, because an answer engine usually cites the URL search kept, not the passage sitting on the URL that was folded away.

• Keep this status separate from a sitewide canonical mistake.
• Record both the user-declared canonical and the Google-selected canonical.
• Treat the selected URL as the one search, and often an answer engine, will reuse.

Find which signal overrode the tag

Google chooses a canonical from more than the link element. Internal links, the XML sitemap, redirects, hreflang, and which URL is stronger can all outweigh the tag. A common pattern is a self-referencing canonical on a URL that the menu still links as a different host, trailing slash, or parameter. Another is an English page and a Traditional Chinese page that repeat the same sentences, each claiming to be canonical, until Google collapses them into one.

Inspect one example and write the two URLs side by side. If they are the same document, change the losing URL so its canonical, its internal links, and the sitemap all name the URL you want kept. If they answer different questions — a service in Central versus a guide about canonicals, or an English page and a zh-HK page with different examples — they are not duplicates. Give each a distinct title, a distinct first answer, and a self-referencing canonical, and stop asking search to merge them.

Point every hint at the URL you want cited

After you pick the URL, make the other signals consistent. Link to it from a page Google already crawls. List that URL in the sitemap, and take the losing copy out. Redirect exact copies that do not need their own address. Leave the canonical self-referencing on the winner and pointing at the winner from the variants. On a Hong Kong site, decide the host, the trailing slash, and whether English and Traditional Chinese are separate documents before you repeat the tag across templates.

The GEO consequence is the same URL. A citation-ready passage on an address Google has folded into another page will not travel. Put the bounded answer on the canonical you want kept, with the organisation and the place named beside the claim, then request indexing for that URL only. Watch the same template until the user-declared canonical and the Google-selected canonical match. That agreement is the fix. Another request on the URL that lost only repeats a hint search has already declined.

• Align internal links, the sitemap, and the canonical on one URL.
• Keep real language versions separate, and consolidate only true copies.
• Re-check URL inspection until the declared and selected canonicals are the same.

What does Duplicate, Google chose different canonical than user mean?

Google saw your canonical tag and indexed a different URL as the main copy. The tag is a preference. Internal links, the sitemap, redirects, and the stronger page can override it.

Should you request indexing again for the URL that lost?

No. Request indexing for the URL you want kept, after the canonical, the internal links, and the sitemap all name it. Repeating the request on the losing URL asks Google to ignore a choice it has already made.