Duplicate Content Checker

What the duplicate Content Checker does

Duplicate content is rarely plagiarism. Far more often it is two of your own pages saying the same thing — a service page rewritten for a second city, two product variants sharing a description, a category page and its top article covering identical ground, or content syndicated to a partner and republished verbatim.

This checker compares two blocks of text and reports overlap two ways. The overlap figure is the share of Text A's words that fall inside a phrase also present in Text B, using a sliding window of the phrase length you choose. That is the number that matters for duplicate detection, because it measures shared passages rather than shared vocabulary. The overall similarity figure compares the unique words in each text and gives a looser sense of whether the two are about the same subject.

The phrase-length setting controls sensitivity. Four words catches paraphrasing and common constructions, and produces false positives on ordinary English. Six is the balanced default and is roughly the point at which a shared phrase stops being coincidence. Eight or ten will only fire on genuinely copied sentences.

What to do about a high score depends on why. If both pages should exist, differentiate them — different examples, different emphasis, genuinely different information. If one is a duplicate of the other, canonicalise it to the original or redirect it. If the shared text is boilerplate such as a shipping policy repeated across product pages, that is normal and search engines handle it routinely; the question is only whether the unique part of each page is substantial enough to justify indexing it separately.

How to use it

1. Paste the two texts you want to compare into the boxes.
2. Leave the phrase length at six words unless you are specifically hunting for paraphrasing.
3. Read the overlap percentage, then look at the highlighted view to see where the shared passages actually are.
4. Decide per page: differentiate, canonicalise, or leave it if the overlap is boilerplate.

A worked example

Two service pages for different cities:

Overlap of A found in B    82.4%   Near-identical
Shared phrases              37
Longest shared passage      64 words

That is the classic doorway-page pattern: one template with the city name swapped. Both pages will struggle, and Google may index only one. The fix is not to spin synonyms — it is to put something genuinely local on each page: the actual team who covers that area, real projects, local pricing, travel times.

Compare that with two pages sharing a 40-word shipping policy inside 900 words of unique copy. That reports around 5% overlap, and it is entirely normal. Boilerplate repeated across a site is not a duplicate content problem; a page that is mostly boilerplate is.

Frequently asked questions