Duplicate without user-selected canonical: what it means, and the fix

What the status means

Google’s help describes it in three parts:[1]

• The page is a duplicate. Google sees it as a copy of another page.[1]
• It names no canonical. The page does not say which address is the preferred one.[1]
• Google chose for you. Google picked the other page as the canonical, and so will not serve this page in Search.[1]

Google adds: “This is not an error, but is working as intended, because Google does not serve duplicate pages.”[1]

Why a site has duplicate pages

Google lists five reasons a site may have duplicate content:[2]

• Region variants. Separate addresses for two countries with essentially the same content in the same language.[2]
• Device variants. A page with both a mobile and a desktop version.[2]
• Protocol variants. The HTTP and HTTPS versions of a site.[2]
• Site functions. The results of sorting and filtering a category page.[2]
• Accidental variants. A demo version of the site left open to crawlers.[2]

See which page Google chose

1. Open the Page indexing report in Search Console. In the table of reasons pages are not indexed, click this status to see the addresses it covers.[1]
2. Inspect one of them. Google says to inspect the address to see which one it considers canonical for the page.[1]
3. Read “Google-selected canonical” under Page indexing in the inspection result.[4] Only the indexed result shows it: the live test cannot predict which version will be canonical.[4]
4. Compare the two pages. Our advice: open both addresses in your browser and decide which case below is yours.

How to fix it: three cases

• Google chose the page you wanted. Nothing is broken. Google calls the status working as intended.[1]
• Google chose the wrong page. Google says you can explicitly mark the canonical for the page.[1] The ways to do that are in the next section.
• The page is not a duplicate. Google says to make sure the content differs substantially between the two pages.[1]

Ways to name the canonical

Google lists three, in order of how strongly each can influence its choice:[3]

1. A redirect. A strong signal that the target of the redirect should become canonical.[3]
2. A rel="canonical" link. A strong signal that the address it names should become canonical.[3]
3. The sitemap. A weak signal that helps the addresses in a sitemap become canonical.[3]

The methods stack: Google says using two or more raises the chance that your preferred address appears in search results.[3] It also says a canonical preference is a hint, not a rule: Google may choose a different page than you do.[2]

What not to do

• Do not use robots.txt to pick a canonical. Google may still index an address that robots.txt disallows, without its content.[3]
• Do not use the URL removal tool for this. It hides every version of the address from Search.[3]
• Do not name two canonicals for one page, for example one address in the sitemap and another in rel="canonical".[3]
• Do not use noindex to steer the choice within one site. Google does not recommend it, because it blocks the page from Search completely.[3]
• Do put a rel="canonical" link on the canonical page itself.[3]

Two statuses that look alike

• Alternate page with proper canonical tag. The page points at its canonical, and the canonical is indexed. Google says there is nothing you need to do.[1]
• Duplicate, Google chose different canonical than user. You marked the page as canonical, and Google thinks another address makes a better one. Google has indexed the one it chose.[1]

Where Convertlyft helps

The free SEO audit crawls up to 25 pages: it reads robots.txt and the sitemap and audits each page it opens. Which page Google chose as the canonical only Google can say: read it in URL Inspection. Connect Search Console to Convertlyft and Google’s clicks and times shown sit beside your visits; the connection is read-only, unless you choose to submit your sitemap to Google from the SEO room. Related: how to check if Google has indexed your site, and the status Discovered - currently not indexed.