Duplicate content: how to find it and fix it

What is duplicate content?

The term “duplicate content” means content that exists in more than one copy. We speak of duplicate content when identical or very similar content is accessible at several addresses.

There are two cases. The exact duplicate: the same HTML, for example a page accessible with and without a tracking parameter. The near-duplicate: the same main text with small variations. That is the case with two product pages that differ only by color.

Most of the time, duplication is technical and unintentional. The CMS, filters, URL parameters or regional versions multiply the addresses of a single page. Nobody decided it.

Does duplicate content lead to a Google penalty?

No, as long as there is no intent to manipulate. Google says so in its page on canonicalization (updated on August 20, 2026). Some duplicate content on a site is normal and does not violate its spam policies.

Its guide to how Search works describes what happens. During indexing, Google groups similar pages and chooses a canonical URL, the most representative of the group. That is the one that appears in the results. The others remain known but are rarely shown.

Duplicate content therefore has a cost, without any penalty:

• The wrong URL displayed: Google may choose a version with parameters or a sorting page.
• Scattered signals: the links received are split between several addresses instead of adding up.
• Wasted crawling: the bot spends time on copies instead of new pages.
• Skewed tracking: traffic and positions split across several URLs in your reports.

The line is clear. Google’s spam policies (updated on August 28, 2026) target “scraping”. It consists of taking content from other sites, often in an automated way, to manipulate rankings. There, the risk of a penalty is real.

Where does duplicate content on a site come from?

Google cites five families of causes: regional variants, device variants, protocol variants, site functions and accidental variants. In practice, they take these forms:

Cause	Example	Usual treatment
Protocol and host	http:// and https://, with and without www	301 redirect to a single version
Trailing slash, capital letters	/page and /page/, /Page/	301 redirect and consistent internal links
URL parameters	?utm_source=, ?sort=price, ?session=	Canonical tag pointing to the URL without parameters
Filters and sorting	Category filtered by color or price	Canonical or noindex depending on search demand
CMS taxonomies	Tags, archives by date or by author	Noindex if no value of their own
Regional versions	fr-FR and fr-BE with identical text	hreflang between the versions
Reused texts	Supplier descriptions, republished articles	Rewriting or canonical to the original
Test environments	Staging site accessible and indexed	Password protection

For regional versions, Google adds a clarification in its documentation on localized versions (updated on September 21, 2026). They only count as duplicates if the main content is not translated. Pages in the same language for two countries must therefore be linked with hreflang annotations.

How do you detect duplicate content?

1. Read Search Console: in URL Inspection, compare “User-declared canonical” and “Google-selected canonical”. The indexing report also lists “Duplicate” pages. See our article on Google indexing.
2. Crawl the site with a tool such as Screaming Frog. According to its tutorial on duplicates, it detects exact duplicates using an MD5 hash of the HTML. Near-duplicates go through the minhash algorithm, with a similarity threshold set to 90% by default.
3. Compare the tags: identical titles, meta descriptions and H1s across several URLs often signal pages that are too similar.
4. Search for an exact sentence in quotation marks on Google to find external copies of your texts.

Which solution should you choose against duplicate content?

Google ranks canonicalization signals by strength in its page How to specify a canonical URL (updated on July 10, 2026). A permanent redirect and the rel="canonical" tag are strong signals. Inclusion in the sitemap is a weak signal.

Solution	When to use it	Effect
301 redirect	The variant does not need to remain accessible	Strong signal, the user lands on the right page
Canonical tag	The variant must remain visible (sorting, tracking, print)	Strong signal, a hint rather than a directive
Sitemap	List only canonical URLs	Weak signal, as a complement
Noindex	Page useful to visitors but with no search value	Removes the page from the index
Hreflang	Language or regional versions	Associates the variants instead of pitting them against each other
Merging or rewriting	Two pages on the same intent	Removes the cause

The canonical tag is defined by IETF RFC 6596 (April 2012). It goes in the head of each variant:

<link rel="canonical" href="https://www.example.com/running-shoes/">

Each indexable page also carries a canonical pointing to itself. It neutralizes parameters added by external links. The implementation rules are detailed in our guide to the canonical URL.

What should you do if another site copies your content?

Start by checking that your page is indexed and that it appears before the copy for an exact sentence. If so, the impact is often small.

• Syndication partner: ask for a canonical tag pointing to your original article, or a noindex on the copy.
• Unauthorized copy: contact the publisher, then the host. As a last resort, file a copyright removal request with Google.
• Massive, automated copying: it counts as scraping under Google’s spam policies. You can report it.

Handling duplicate content is part of technical SEO, a workstream presented in our SEO guide. The best prevention remains editorial: one page per intent and texts specific to each page.