Why is Google indexing the wrong page?

Inspect the URL you wanted indexed in Search Console. If the verdict says the page is a duplicate and names a different Google-selected canonical, this is canonicalisation rather than an indexing failure.

The report distinguishes the canonical you declared from the one Google chose, side by side. That pair is the whole diagnosis.

"Duplicate, Google chose different canonical than user" is the exact label to look for in the Page indexing report.

If instead the verdict is blocked, excluded by noindex, or crawled and not indexed, this page is not your answer — those are different problems with different causes.

Internal links. The URL your own navigation, breadcrumbs and body links point at is the strongest statement you make about which address is real.

Redirects. A 301 is an instruction rather than a hint, and it beats a canonical tag pointing the other way every time.

The sitemap. A URL listed there is advertised as canonical; listing the variant you do not want is a direct contradiction.

hreflang annotations, which must be reciprocal and must point at canonical URLs to be used at all.

External links, which you do not control and which can be enough on their own when the internal signals are ambiguous.

Content similarity itself. When two pages say substantially the same thing, they enter the same cluster whether or not you intended them to be alternatives.

Parameterised URLs: tracking parameters, sort orders and session identifiers create variants of one page, and any of them can win if links point there.

Host variants: the apex and `www` serving the same content without one redirecting to the other, or HTTP and HTTPS both answering.

Trailing slash inconsistency, where `/page` and `/page/` both return 200.

Pagination and filtered listings whose content overlaps almost entirely with the unfiltered page.

Syndicated or near-duplicate content, where another page — sometimes on another site — is judged the better representative.

Make every internal link use it. This is the single most effective change and the one most often skipped, because it means editing templates rather than adding a tag.

Redirect the variants with 301 where only one address should exist at all. A redirect removes the ambiguity instead of arguing about it.

List only the preferred URL in the sitemap, and remove the variants.

Keep the canonical tag self-referential on the preferred page and pointing at it from the variants.

Make all four agree. A canonical tag contradicted by the other three signals is a hint being outvoted, which is exactly what is happening now.

Google has to re-crawl the pages involved before anything changes, and the re-crawl schedule is not something you set.

Requesting indexing in URL Inspection queues one URL; it does not guarantee a timeframe and it does not scale past a handful.

Canonicalisation is a cluster decision, so the change takes effect when enough of the cluster has been re-crawled, not when the first page has.

Weeks is a normal duration on a site crawled infrequently. Changing the signals again halfway through restarts the process.

Watch the Google-selected canonical field in URL Inspection rather than the search results: it flips before the visible result does, which is the earliest confirmation the change landed.

The contradiction is mechanically visible without waiting for Google. Every URL that returns 200 and declares a canonical pointing elsewhere is a variant that should probably not be reachable.

VeriFixScan reports `seo.canonical` per page, which surfaces pages whose declared canonical is not themselves.

`structure.url_consistency` reports trailing-slash and path inconsistencies across the crawled set, and `domain.url_variants` reports host variants answering independently.

`structure.duplicate_content` reports pages whose bodies are substantially the same, which is the cluster Google is choosing from.

Together they name the variants competing with your preferred URL, which is the list to redirect or de-link.