What Is a Good Similarity Score?
There is no universal pass mark for a similarity score. Learn what the percentage measures, what affects it, and how to judge a report by its matches.
01
What the percentage actually measures
A similarity score reports how much of a checked document matched text found in the sources that were searched. It does not measure intent, originality of ideas, or whether the writing was produced with AI assistance. Two documents with the same score can have completely different problems.
That is why a single target number is a weak goal. A report is useful when it points you to the passages that need a citation, a quotation, or a genuine rewrite.
02
What moves the number
Several ordinary parts of academic writing create matches: correctly quoted material, reference lists, titles and headings, standard methodology language, and assignment instructions that appear in many submissions. Longer documents also have more opportunity to accumulate matches.
Source coverage matters too. A check only reports what it searched, so a low score does not guarantee that every possible source has been considered.
03
How to judge your own report
Read the highlighted passages first and ask whether each one is attributed. A match inside quotation marks with a citation is usually doing its job. A close paraphrase without a citation usually needs attention.
If your institution publishes a threshold or a review policy, follow it — but treat the policy as a starting point for human review rather than a substitute for it.

