Keyword Density Checker
Analyse text to find the most-used words and phrases and their density — with the honest position that there is no optimal density to hit.
0 words
Only 4 words in English reach 3%
The standard advice is to aim for a keyword density somewhere around 2–3%. It is worth seeing what that number means. Measured against 339,742 words of ordinary English prose — the FAQ text on this site, written by a person rather than sampled from a table — here is what words actually reach those densities:
| Rank | Word | Share of all text | |
|---|---|---|---|
| 1 | the | 6.05% | |
| 2 | a | 4.14% | |
| 3 | is | 3.51% | |
| 4 | and | 3.08% | |
| 5 | it | 1.74% | |
| 10 | than | 0.86% | |
| 50 | has | 0.23% | |
| 100 | wrong | 0.11% |
Exactly 4 words in the entire corpus reach 3%, and only 8 reach even 1%. So “aim for 3% keyword density” is asking you to use your keyword as often as the word and, and more often than it. Text that does that does not read like text — which is roughly what a search engine concluded twenty years ago.
Because word frequency is a power law
Words have never been spread evenly. In the same corpus:
| The most common… | account for |
|---|---|
| 10 words | 24% of everything written |
| 100 words | 50% of everything written |
Half of all the text you have ever read is a hundred words, over and over. That is roughly Zipf's law — frequency multiplied by rank stays in the same neighbourhood as you go down the list. In this corpus that product runs from about 20,600 at rank 1 to about 44,500 at rank 1,000: the same order of magnitude across three orders of magnitude of rank, though drifting upward rather than holding perfectly flat, which is the usual finding in real text and means the exponent is close to but not exactly 1.
The practical consequence: a density figure is dominated by how common your word already is in English, not by how thoroughly you covered the topic. “Bank” and “photosynthesis” cannot reach the same density in readable prose no matter what either page is about. Use the number to spot text that has gone strange — the same phrase eight times in a paragraph — rather than as a target to hit.
How to use
- Paste your page text.
- Read the ranked words and phrases.
- Look for accidental repetition rather than a target number.
- Write for readers and check the density afterwards.
Frequently asked questions
What is the ideal keyword density?
There isn't one, and any specific figure you have seen is folklore. Search engines moved well beyond counting term frequency years ago, and Google has said directly that there is no target density. The number is diagnostic, not a goal.
Then what is this actually useful for?
Spotting accidental repetition, checking that your page is obviously about what you intended, and finding filler. A page whose top terms are not its subject is worth rewriting — which is a genuine insight the count can give you.
Is keyword stuffing penalised?
Yes, explicitly. Repeating a term unnaturally is a documented spam practice and can harm a page's performance as well as reading badly. The threshold is not a number — it is whether the text reads as written for a person.
What matters more than density?
Covering the subject properly, matching what someone searching would actually want, and being genuinely useful. Modern retrieval works on meaning rather than exact term matching, so a page that answers the question well outperforms one engineered around a phrase.
Should I include synonyms and related terms?
Naturally, yes — writing about a subject properly produces them without effort. Deliberately sprinkling variations to hit some notion of semantic coverage is the same mistake as stuffing, one abstraction layer up.
Does this count stop words?
It can, and filtering them is usually what you want. Otherwise 'the' and 'and' dominate every result and tell you nothing. Their presence at the top is a sign the filter is off rather than a finding about your writing.
🔒 This tool runs entirely in your browser. Nothing you enter is uploaded, logged, or stored.