AdSense rejected for duplicate content — how to fix it
Two different things arrive under this name. Either you republished material from elsewhere, or — far more often — your own site serves the same content at several addresses. The second case is a configuration problem, it is invisible from the front end, and it is the one most publishers never think to check.
How to fix it
- Find how many URLs serve the same page. http and https, www and non-www, with and without a trailing slash, with tracking parameters appended, and printer-friendly or AMP variants. Each combination that returns 200 is another copy of the same article.
- Pick one canonical hostname and redirect the rest. A single 301 to the chosen form, applied at the server, collapses most of the duplication in one change.
- Set a self-referencing canonical tag on every page. Every page should declare its own preferred URL, so a version reached with a tracking parameter still points at the clean one.
- Deal with archive and pagination duplication. Tag, category, date and author archives can each list the same excerpts. Keep the ones with a purpose and noindex the rest.
- Remove or rewrite anything republished from elsewhere. Manufacturer descriptions, press releases, syndicated articles and affiliate feed copy are all duplicate content even when you are permitted to use them.
Two rejections wearing one name
When AdSense declines a site for duplicate content, one of two quite different
things has happened.
The first is what everyone assumes: material on the site is republished from
somewhere else. Manufacturer descriptions, press releases, an article
syndicated from a partner, or content lifted outright.
The second is far more common among publishers who write everything themselves:
the site is duplicating itself. The articles are original, and every one of
them is served at four or five different addresses. From a crawler's perspective
that is not one article, it is five identical ones.
How a site ends up serving five copies of everything
A default install commonly answers all of these with HTTP 200:
http://example.com/article
http://www.example.com/article
https://example.com/article
https://www.example.com/article
https://example.com/article/
Add a tracking parameter and you have another. Add a printer-friendly view, an
AMP variant, or a feed that publishes full posts, and the count climbs again.
None of this is visible from the front end. The site looks completely normal to
its owner, because a browser follows whichever link was clicked and never shows
that four other addresses would have worked.
The fix is two changes, not a content rewrite
Redirect to one canonical form. Choose https and either www or non-www, then
301 everything else to it at the server. This is a single configuration change
and it collapses most of the duplication immediately.
Declare a self-referencing canonical on every page. Each page states its own
preferred URL, so a visit that arrived with a tracking parameter still resolves
to the clean version. Most CMS platforms and SEO plugins do this by default —
check that yours actually is, rather than assuming.
Archives are the leftover case
Once URLs are consolidated, what usually remains is archive duplication. A site
with an enthusiastic tagging habit produces a tag archive, a category archive, a
date archive and an author archive that all list the same excerpts in slightly
different orders.
The test is whether a human would use it to find something. A category archive on
a site with five categories, yes. A tag archive for a tag applied to one post,
no — noindex it. That also removes it from the pool of thin monetised pages that
cause a separate policy problem.
Republished material has no version that works
If pages are built on supplied copy — manufacturer descriptions, feed data, press
releases — permission to use the text does not make the page reviewable. The same
paragraphs are on every other site selling the same thing, so the page adds
nothing that is not already published.
The only fix is to add what is not in the feed: your own photographs, your own
measurements, what the thing was actually like to use. That is more work than
rewording, and it is the only version that survives review.
Frequently asked
- I never copied anything. Why was I rejected for duplicate content?
- Almost certainly because your own site serves each article at more than one address. A default CMS install commonly answers on http and https, with and without www, with and without a trailing slash, and again for every tracking parameter. From a crawler's point of view those are separate pages with identical content.
- Does using manufacturer product descriptions count?
- Yes. Being licensed to use text has no bearing on whether it is duplicate — the same paragraphs appear on hundreds of other sites, so the page adds nothing that is not already published. Product pages built on supplied copy need your own testing, photographs or judgement to become reviewable content.
- Are tag and category archives duplicate content?
- They can be. A site with a tag for every noun produces dozens of archives listing the same excerpts in different orders. Keep the archives that genuinely help someone navigate and noindex the rest.
- How is this different from scraped content?
- Duplicate content is usually accidental and structural. Scraped content is deliberately taking someone else's work. Google treats the second far more harshly, and it is much harder to recover from.
Check whether this is still blocking you
The free Monetific checker tests your site against this and every other mechanical requirement, and tells you what is still outstanding before you re-apply.
Run the free check