Draftcamp

Keyword density checker

Word and phrase frequency, with a warning when repetition tips into stuffing

Runs entirely in your browser - your text is never uploaded

Nothing leaves your browser

Paste content to see its vocabulary

Single words, pairs and three-word phrases, ranked by how often they appear.

Keyword density is a 2005 metric

There is no density target. Google has said, more than once and unambiguously, that there is no ideal keyword density and no ratio that triggers a penalty. Every "aim for 1 to 2%" article is repeating advice that stopped being true before most of the people quoting it started working in search.

So why measure it at all? Because the number is still diagnostic, just not in the way the old advice claimed. Density does not tell you whether Google will rank the page. It tells you what you actually wrote about, which is often not what you meant to write about.

What to look for

  • A term you did not intend at the top. If your page about content audits is topped by "team", the emphasis drifted.
  • Your actual topic missing from the top ten. The clearest sign a page is about something adjacent to its title.
  • The same phrase twelve times. Almost always accidental, and almost always makes at least three sentences clumsy.

The better question

Not "how often does this phrase appear" but "does this page answer the question". That is what an answer engine evaluates, and it is what the readability checker gets closer to by looking at structure rather than frequency.

Common questions

What is the ideal keyword density?
There is not one. Google has repeatedly said there is no keyword density target and no threshold that triggers a penalty. Anyone quoting "1 to 2%" is repeating advice from roughly 2005. What density is genuinely useful for is spotting unintentional repetition - a phrase you leaned on twelve times without noticing.
Can keyword stuffing still hurt a page?
Yes, but not because of a ratio. Stuffing hurts because the writing gets worse: repeated phrases make sentences awkward, readers bounce, and answer engines find nothing clean to quote. The damage is to the reader, and the ranking follows the reader.
Why are two-letter words missing?
The tokenizer drops tokens of two characters or fewer, which is inherited from the same tokenizer our page scorer uses. It means "AI" and "UX" do not appear. That is a deliberate trade: one tokenizer across every tool, so two tools never disagree about what a word is.
Why do some phrases not appear?
Phrases made up entirely of stopwords - "of the and" - are dropped, because they crowd out anything meaningful. A phrase keeps its stopwords as long as it contains at least one content word, so "how to write" survives.