Keyword Density Checker
Word frequency and two- and three-word phrases for a block of copy, with stop words filtered out.
Nothing you type here is uploaded, stored or sent anywhere. It all runs in your browser.
Density is a symptom, not a target
There is no density figure Google rewards. Nobody has ranked because they hit 2%. The number is useful in one direction only: as a warning that copy has drifted into repeating itself.
- Above roughly 4% on a single term, writing usually reads badly. Read it aloud — if it grates, it is too much.
- Phrases matter more than words. The two- and three-word view tells you what the page is actually about.
- Variation beats repetition. Search engines have handled synonyms for years; your reader has always preferred them.
Questions people ask
If density is not a ranking factor, what is the number for?
Editing. The figure is useful as a diff — run a draft, run it again after a rewrite, and see whether the phrase you were leaning on has stopped dominating. It is also the fastest way to catch generated copy, which loops on the same three-word phrase far more than a person would.
What density should I aim for?
None. Every tool that prints a target range inherited it from the early 2000s, when term frequency really was most of the algorithm. The only number worth reacting to is a high one, and only because it usually means the writing has gone stiff.
Does it matter more where the keyword appears than how often?
Much more. The term in the title, in the H1, in the first hundred words and in one subheading does the work. The eleventh mention in paragraph nine does nothing the third did not already do.
Should I match the density of the page currently ranking first?
No. Their density is an output of how they wrote, and copying an output does not give you the input. If you want to learn from a competing page, read what it covers that yours does not — that comparison pays; the percentage does not.
Is keyword stuffing still something to worry about?
It is still named in the spam policies, so yes in principle. In practice the realistic outcome is not a manual action — it is that the page reads badly, nobody quotes it, nobody links to it, and it loses to something written for people. The punishment arrives as indifference rather than a penalty.
Can I run a client draft through this before it is published?
Yes. Nothing you paste leaves the page — the analysis runs in your browser and there is no server behind it, so unpublished copy stays unpublished.
Guides for this tool
Related tools
Copy is on point but rankings are not
Once the page says the right things, the gap is usually authority.