It means Google found two or more pages on your site that look the same or nearly the same, you didn't say which one is the main version, so Google picked one itself and indexed that instead. Google describes this status as normal behaviour, not an error. You only need to act if Google picked the wrong page, or if the pages aren't really duplicates and you want both to appear in search.
What Google says this status means
Search Console's Page indexing report help (checked 13 November 2026) explains that the page is a duplicate of another page, that it doesn't declare a preferred version, and that Google has chosen the other page to show in Search. Google says it “is not an error, but is working as intended”, because it doesn't serve duplicate pages.
Google's guide to what canonicalisation is (last updated 20 August 2026, checked 13 November 2026) fills in the background. When Google finds pages with the same or very similar main content, it groups them and chooses the most complete and useful one as the “canonical”: the version it indexes and normally shows. It also says some duplicate content on a site is normal and isn't against its spam policies.
So this isn't a penalty. It's Google tidying up.
Where duplicates come from on small business sites
Most of the time, nobody created a duplicate on purpose. The usual causes:
| Cause | Example |
|---|---|
| Tracking or ad parameters | /services?utm_source=newsletter and /services |
| Sorting and filters | /cakes?sort=price and /cakes |
| www and non-www, http and https | http://example.com/about and https://www.example.com/about |
| Trailing slash differences | /contact and /contact/ |
| Print or “share” versions | /menu/print alongside /menu |
| Tag, category or archive pages | A tag page that lists one post and repeats its text |
| A staging or test copy left open | staging.example.com showing the same site |
| Near-identical service or area pages | Two pages that differ only by one place name |
Google's canonicalisation guide lists several of these, including protocol variants, sorting and filtering, and test versions accidentally left open to crawlers.
Step 1: Find out which page Google picked
Open URL Inspection in Search Console and paste in one of the affected URLs. Under Page indexing, look at Google-selected canonical. That's the version Google indexed.
One thing that catches people out: the Test live URL button doesn't check duplicate or canonical conditions, according to the same help page. So the indexed result is the one that matters here, not the live test.
Step 2: Decide whether Google got it right
Ask one question for each example URL: is the page Google chose the one I'd want customers to land on?
- Yes, Google chose well. The duplicate is a parameter version, a print view or an old http address, and the clean page is indexed. You can leave it. Adding a self-referencing canonical tag to your main pages is good housekeeping, but it isn't urgent.
- No, Google chose the wrong page. For example, it indexed
/services?ref=footerinstead of/services, or an old page instead of the new one. Go to Step 3. - These aren't really duplicates. You meant them to be separate pages, but they're too similar for Google to tell apart. Go to Step 4.
Step 3: Tell Google which page you prefer
Google's guide on how to specify a canonical (last updated 10 July 2026, checked 13 November 2026) ranks the signals by strength:
- Redirects are a strong signal. Use a permanent redirect when the duplicate shouldn't exist at all, such as an old address or an http version.
- A
rel="canonical"link is also a strong signal. Use it when both URLs need to keep working, such as a filtered view of a list. - Including the URL in your sitemap is a weak signal, but it helps.
Signals add up, so use more than one where you can. Google also recommends linking to your preferred URL consistently from your own pages.
And three things not to do, all from Google's guide:
- Don't use robots.txt to handle duplicates.
- Don't use the URL removal tool for this. It hides every version.
- Don't send mixed signals, such as one URL in your sitemap and a different one in the canonical tag.
Google also advises against using noindex to choose between duplicates on your own site, because it blocks the page from Search completely.
If you use WordPress, Wix, Shopify or another platform, you usually set canonicals through its SEO settings or a plugin rather than by editing code. If you're not sure, ask whoever manages your site.
Step 4: If the pages should be separate, make them genuinely different
If Google has grouped two pages you want indexed separately, the fix isn't technical. Google's troubleshooting guide (last updated 21 August 2026, checked 13 November 2026) says fixing this comes down to making sure clustered pages are sufficiently different, and that pages split apart faster when the difference is clear and significant.
For two similar service pages, that means real differences: what each service includes, who it's for, how it's priced, how long it takes, and examples or photos specific to each. Swapping a few words won't do it. If you can't find real differences, they're probably one service, and one stronger page will serve customers better than two thin ones.
A hypothetical example
This is an invented example, not a client result. A home bakery's site lists cakes at /cakes, with buttons to sort by price or by flavour. Search Console shows /cakes?sort=price and /cakes?sort=flavour as “Duplicate without user-selected canonical”. URL Inspection shows Google picked /cakes as the canonical.
That's the right choice, so nothing needs fixing. If the bakery wants to make its preference explicit, it can add a canonical tag on all three URLs pointing to /cakes, and make sure the menu links to /cakes without parameters.
How it differs from the look-alike statuses
| Status | What it means | Usually needs action? |
|---|---|---|
| Duplicate without user-selected canonical | You didn't name a main version, so Google picked one | Only if Google picked wrong |
| Duplicate, Google chose different canonical than user | You named a main version, but Google preferred another | Check why your choice was overruled |
| Alternate page with proper canonical tag | The page correctly points to its main version | No |
| Crawled – currently not indexed | Google visited and chose not to index, for now | Only for important pages |
If you're seeing the last one, our guide to “Crawled – currently not indexed” covers it.
How long changes take
Not instant. Google's troubleshooting guide says it may keep pages in a duplicate group for up to two weeks even after you fix the content, and Search Console reports update only after Google recrawls. Make your change, request indexing for the main URL in URL Inspection, and check back in a few weeks rather than a few days.
What's in your control, and what isn't
In your control: which URLs exist, where your internal links point, whether you use redirects and canonical tags consistently, and whether pages you want indexed separately are genuinely different.
Not in your control: which canonical Google finally chooses. Google calls a canonical preference a hint, not a rule, and indexing is never guaranteed. Clear signals make your preference much more likely to stick; they don't force it.