Short answer: Duplicate content is the same or very similar content at more than one address. Google says some of it is normal and isn’t a spam violation: it groups the copies and shows one, the canonical. The trouble starts with content copied from other sites and with near-identical pages built to rank, such as one per city with the same text.
Which duplicates are a problem
| Situation | A problem? | What Google says |
|---|---|---|
| HTTP and HTTPS versions of a page | No | A normal duplicate it groups under one canonical |
| Sorted and filtered category pages | No | A normal duplicate from site functions |
| A page translated into another language | No | Not a duplicate when the main content is translated |
| Content copied from other sites | Yes | Scraping, against its spam policies |
| Near-identical pages for each city | It can be | Doorway abuse if substantially similar or if they funnel visitors to one page |
From Google Search Central, “What is canonicalization” and “Spam policies for Google web search”, read November 3, 2026.
How Google handles copies of a page
When pages look the same, or their main content is very similar, Google clusters them and picks the one it finds most complete and useful as the canonical. It crawls that page most often and the others less.
The canonical is also the main source Google uses to judge content and quality. You can state your preference with redirects, your sitemap or a canonical tag, but Google treats it as a hint. Our guide to the canonical tag covers each method.
When duplication becomes spam
In Google’s spam policies, scraping is republishing other sites’ content without adding original content or value. Scaled content abuse is producing many unoriginal pages to manipulate rankings, including by translating or rewording copied content.
Doorway abuse covers pages aimed at specific regions or cities that funnel visitors to one page, and substantially similar pages that look more like search results than a clear site. A bilingual site is fine when each version is a real translation, which our guide to hreflang helps you mark.
How to keep your content unique
- Choose one address for each page and point the copies to it.
- Send HTTP, www and capitalized variants to that address with a 301 redirect.
- Translate the main content of each language version, not only the menus.
- Give each city or service page facts of its own, not the same text with a new name.
- Don’t republish other sites’ content without adding something of your own.
- Compare a new page with your closest pages before you publish it.
- Merge, redirect or rewrite the weaker page when two cover the same thing.
Our guide to the 301 redirect explains how to point an old page to the one you keep.
How we check our own pages
Before publishing, we compare every new guide with its closest pages in English and Spanish, counting the 5-word and 8-word sequences they share. For the eight guides we wrote on November 3, that was 128 comparisons, and the highest share was 2.7%.
The first draft of our structured data guide shared 5.6% with our guide to local business markup. We rewrote it until it shared less than 3%.
At 3:49 AM we also measured our eight SEO city pages against each other. They share up to 12% of their 8-word sequences in English and 13.3% in Spanish. The longest shared passages are the booking box, the source lines and the links to the other cities.
Before you publish a page
- The page has one address, and its copies point to it.
- HTTP, www and capitalized variants lead to the same page.
- Each language version has its own translated body.
- City and service pages carry facts of their own.
- Nothing is republished from other sites without your own value.
- The page shares little text with your closest pages.
RememberSome duplicate content is normal, and Google picks one version for you. Copied content and near-identical pages made to rank are the real risk, so give every page something of its own.
If you want us to find the duplicates on your site, see our technical SEO service.
Quick answers about duplicate content
Is there a duplicate content penalty?
Google says some duplicate content on a site is normal and isn’t a violation of its spam policies: it groups the copies and shows one. Copying other sites or making near-identical pages to rank is a different matter.
Are translated pages duplicate content?
Not when the main content is translated. Google treats language versions as duplicates only if the body stays in the same language and only the header, footer or other minor text changes.
Can I use the same text on several city pages?
Google’s spam policies list pages aimed at specific regions or cities that funnel users to one page, and substantially similar pages, as doorway abuse. Give each city page facts of its own.
Which version of a duplicate will Google show?
Usually the canonical, the page Google judges the most complete and useful in the group. Redirects, sitemaps and canonical tags are hints, so Google may still choose another page.
Does duplicate content waste crawling?
Google crawls the canonical most often and the duplicates less, to lighten the load on your site. Many copies can still make it harder for you to see how a page performs in search.
Sources
Read on November 3, 2026. Google Search Central: What is canonicalization and Spam policies for Google web search. Our own case: the overlap checks of our guides and our SEO city pages, on November 3, 2026.