TL;DR: Thin, duplicate and overlapping content are three different problems with three different fixes. Duplicate content is the same material at more than one URL: choose one version and point the rest at it. Overlapping content is two of your pages doing the same job: merge them, or give each its own purpose. Thin content is a page that gives a visitor little they could not get elsewhere: improve it, or remove it if it has no purpose. Google does not penalise ordinary duplication, but it picks one version to show and may not pick yours. Copying content from other sites and mass-producing pages to manipulate rankings are different, because those break its spam policies.
It’s important to work out which problem you have before attempting a fix. This guide covers how Google treats each one, how to find them on your own site and what to do next.
Does Google penalise duplicate or thin content?
Not for ordinary duplication. Google said in 2008 that there is no “duplicate content penalty” in the way most people mean it, and its current guidance on canonical URLs describes the same behaviour. When it finds duplicate pages, it groups them, picks the version it thinks is best and consolidates signals such as links onto that one. Its ranking systems also aim to show only the most relevant results when many are very similar.
So the cost is a loss of control, not a punishment. Google may show a version you did not choose. Links to your pages can end up split across several URLs. And Googlebot spends time crawling copies when it could be reading your new pages.
Deliberate copying is a different matter. Google’s spam policies name three practices that matter here. Scraping is republishing content from other sites without adding original value, to manipulate rankings. Scaled content abuse is generating many pages mainly to manipulate rankings, however they are made. Doorway abuse is creating pages to rank for specific, similar queries that act as less useful stepping stones to the page visitors actually want. Google says sites that break these policies may rank lower or not appear in results at all.
Thin content has no neat definition from Google, and it is not a word count. Google lists writing to a particular word count among its warning signs and says it has no preferred one. What it asks instead is whether a page offers original information or analysis, and substantial value compared with other pages in the results. In practice, a short page that fully answers a narrow question can pass, and a long page that repeats what everyone else says can fail. If you have read about the Helpful Content Update, Google now lists it as retired, because it became part of its core ranking systems in March 2024.
AI-written pages are not banned. Google says appropriate use of AI is not against its guidelines, while generating many pages without adding value for users is scaled content abuse. A page produced by AI with nothing original added is simply the thin content problem in a new form.
How do you find thin, duplicate and overlapping content?
Work through these five checks in order. Each one tends to surface a different problem.
- Start with what Google already treats as a duplicate. In Search Console’s Page indexing report, look at the reasons “Duplicate without user-selected canonical” and “Duplicate, Google chose different canonical than user”. The first means Google has chosen a canonical for you, which it says is not an error. The second means Google disagrees with the canonical you declared, which deserves a closer look. More generally, Google says a page marked as a duplicate is usually fine, because it means the canonical page has been found and indexed. You are looking for cases where Google has picked a page you would not have picked.
- Crawl the site and look for repeats. Export the title, H1 and meta description for every page from a crawler such as Screaming Frog and sort them. Pages that share a title or heading are the quickest pointer to duplicates and to overlapping pages.
- See which of your pages Google shows for the same query. In the Search Console Performance report, open the Queries tab, click a query that matters to you, then open the Pages tab. Google’s guide to common tasks sets out those steps. If two or more of your URLs appear for one query, they may be overlapping. Our guide to keyword cannibalisation goes further on this check.
- Search for a distinctive sentence from the page, in quotation marks. Matches on other sites point to copied or republished text. Matches on your own site point to internal duplication, such as the same paragraph reused across service or location pages.
- Read what is left against Google’s own questions. For the pages these checks flag, and your weakest performers, ask whether the page gives original information or analysis, whether it offers substantial value compared with other pages in the results, and whether it looks mass-produced. Google’s helpful content guidance lists all three. Record the answers in a simple content audit sheet. Pages built from copied supplier descriptions are likely to fail this test, and so are near-identical pages that change only a town name. Read the second kind closely against Google’s doorway abuse policy.
What is overlapping content, and why does it matter?
Overlap is the one problem in this post’s title that most guides skip. Two pages overlap when they are written differently but do the same job for the same search. A pricing page and a “how much does it cost” blog post aimed at the same query would be one example.
Overlap is not a spam policy violation in itself, but Google limits what it will show. Its site diversity system generally will not show more than two listings from the same site in the top results, though it may show more when it judges that especially relevant. A third page aimed at the same query is unlikely to win another slot, and it may split links and internal links away from the other two. If Google judges two pages to be near-duplicates, it picks one to show and leaves the other out.
Not every pair is a problem. A beginner’s explainer and a step-by-step guide on the same subject can sit side by side, because they answer different questions. Search intent is the tie-breaker: if searchers want the same thing from both pages, you have overlap.
Wildcat Example: On a cyber security client’s site, ‘penetration testing services’ and ‘what is penetration testing?’ looked like duplicates, but Search Console showed the first earning clicks from buyers comparing providers and the second from people learning the term, so we kept both because each did a different job. The pair we did merge were two service pages, ‘managed security’ and ‘managed cyber security’, which were showing up for the same searches and splitting the clicks between them.
What should you do about each problem?
Match the fix to the problem. Where you have a choice, keep the page that already earns clicks and links.
| What you found | What to do | Why |
|---|---|---|
| The same page at several URLs (tracking parameters, www and non-www, http and https, print versions) | Pick one preferred URL. Redirect the others if they do not need to stay live. If they do, add a rel=canonical link to each pointing at the preferred URL. Link to the preferred URL everywhere on your site. | Google treats redirects and rel=canonical as strong signals for which URL to show, and says consistent internal links help. |
| Two overlapping pages, one clearly stronger | Move anything worth keeping from the weaker page into the stronger one, then 301 redirect the weaker URL to it. | A permanent redirect tells Google the old URL is the worse version, and only one page is left to compete. |
| Two similar pages that serve different needs | Keep both. Give each its own title, H1 and opening that name the different question it answers, and link between them. | If Search Console reports a page as a duplicate and you disagree, its advice is to make sure the content differs substantially. |
| A thin page that should exist | Add what a visitor needs and cannot get from the pages already ranking: original information, clear answers, evidence from your own work. More words alone will not do it. | Google asks for substantial value compared with other pages in the results, and has no preferred word count. |
| A thin page with no job to do | Remove it, or redirect it to the closest page that answers the same question. If visitors need it but searchers do not, add a noindex tag. | A noindex tag keeps the page out of Google Search entirely. |
| Text copied or republished from another site | Rewrite it with something of your own, or take it down. | Republishing other sites’ content without adding original value is what Google’s scraping policy describes. |
Choose the method to suit the job. A redirect retires a page. A rel=canonical link keeps both pages live but names the one you prefer. A noindex tag removes a page from Google Search altogether, so Google does not recommend it for choosing between versions of a page on the same site, where it prefers rel=canonical. Google can only see a noindex tag if it can crawl the page, so do not also block that page in robots.txt. Do not use robots.txt for canonicalisation either: Google may still index a blocked URL without its content. Whichever method you use, keep your signals consistent. Do not name one URL in your sitemap and a different one in a canonical tag. Our posts on canonical tags and 301 redirects cover each in more detail.
Remove only pages that have no job to do. Google says that adding or removing large amounts of content mainly to make a site seem fresh will not help rankings, so remove a page because it has no purpose, not as a ranking tactic.
Wildcat Example: For a roofing client, we discovered that over 300 location pages had been created that contained thin, not very unique content. Initially we audited the site using Screaming Frog and cross-referenced data from GA4 and GSC from the past year to determine how these pages were performing. The data showed us that these pages received few, if any, clicks over the past year.
We could also see that the page content wasn’t adding much value. As a result, we decided to redirect the pages into six main location pages which were optimised to have unique content and case studies. These pages have now begun to perform much better commercially.
Top Tip: Before you merge or delete anything, write one sentence for each page saying what question it answers for the reader. If two pages give the same sentence, merge them. If the sentences differ, keep both and make the titles say so. Then check what each page earns in Search Console and which sites link to it before you redirect, and send the old URL to the page that now answers its question.
What happens when two sites have almost identical content?
Google generally shows one of them for a given search. Its systems aim to show only the most relevant of very similar results, and it has systems to show original content prominently. Google does not publish what decides which copy wins. If you are the one republishing, adding something the original does not have is the best way to give Google a reason to show your version. Google’s own warning to affiliates, in the 2008 post, is that an identical listing is hard to rank unless you add significant value.
How Wildcat Handles Thin, Duplicate and Overlapping Content
When we take on a new SEO account, one of the first things we do is run specialist SEO audits. Our SEO page says that where a site is not ranking well, those audits show us why and shape the roadmap. On the technical side, our technical SEO page says we work with our content experts to change duplicate content or implement canonicals. For pages that need rewriting, our SEO content writing service includes our content specialists overhauling existing content. If you would like a view on where your own site stands, book a free consultation and tell us what you are seeing in Search Console.
