Start with a clear question
A list of frequent words can reveal a repeated introduction, a technical term or a necessary label used throughout instructions. These have different purposes. Read the surrounding sentences before replacing a repeated term. Using synonyms for a single component can make a procedure less precise, even when the frequency table looks more varied.
A phrase needs a different denominator
For one word, each token is a possible match. For a two-word phrase, the tool compares adjacent pairs within each sentence. Four words in one sentence create three pairs, not four. A phrase is not matched across a sentence boundary. The displayed percentage divides the number of matching sequences by the available sequences of the same length.
Worked example
clear text clear text
| Matching “clear text” pairs | 2 |
|---|---|
| Compared two-word sequences | 3 |
| Frequency | 66.67% |
Try an exact phrase search
Enter the word or expression in the search field. Apostrophe styles are normalized, and you can choose whether the comparison distinguishes case. This is token matching, not substring matching: “art” is not counted inside “article”. Punctuation can separate tokens; inspect unusual names or codes before interpreting a count as a linguistic statement.
Use the common-word filter carefully
The filter hides expressions consisting entirely of common words in the selected language. It does not delete those words from the text or from the denominator. Turning the filter off restores them to the table. This helps a reader compare the same draft without silently changing how its percentages were calculated.
Turn the result into an editorial decision
Check whether a repeated passage adds a fact, names an object unambiguously or repeats a promise without helping the reader. Edit the passage, then compare its meaning and frequency with the original. No particular percentage guarantees rankings. A report is a description of the supplied text, not proof of content quality, originality or search demand.