Keyword Density Checker: What the Numbers Mean

A keyword density checker counts how often each word and phrase appears in a piece of text and shows that count as a share of the total word count. It is a repetition detector: a way to see whether a page leans too hard on one phrase, or mentions a phrase you care about only once. Paste the text, choose how long a phrase to analyse, and read the table.

The number itself is not a ranking target. No search engine rewards a specific density, and chasing one produces worse writing. What density is good for is spotting accidents: the same transition phrase in every paragraph, a product name repeated until it reads like a label, or a target phrase that appears once in 2,000 words when the page is supposed to be about it.

#What the tool reports

• Total words — the token count for the text, after the same whitespace rules used by the word counter.
• Unique words — how many different words appear.
• Unique word ratio — unique words as a percentage of the total. A very low ratio usually means heavy repetition; a very high ratio can mean the text is a list of unrelated terms.
• Keywords table — up to 60 rows of words and phrases with their word count, occurrence count and density.
• Top phrases — the highest-count multi-word phrases, which are usually more useful than single words for judging repetition.
• Guidance — automatic flags for phrases above 3.5% density and for phrases that appear twice or more but stay under 0.25%.

Density is calculated as occurrences of the phrase divided by the total word count. A two-word phrase that appears ten times in a 500-word text has a density of 2%, not 4%, because it is the phrase occurrence that is counted, not each word inside it.

#How to check keyword density

1. Open the keyword density checker.
2. Paste the article, product description or page copy you want to inspect.
3. Set the minimum word length. Keep it at 1 for a raw view, or raise it to 3 or 4 to ignore short filler words.
4. Choose the longest phrase. Single words are quick; two to four words are where repeated phrases become visible.
5. Leave Ignore common stop words on unless you specifically want to count the, and, of and similar words.
6. Press Run and read the table, then look at the top phrases below it.

The analysis runs in your browser. Nothing is uploaded, which matters if you are checking a client's unpublished page or a competitor's draft you were given.

#Single words versus phrases

Single-word counts are easy to read but easy to mislead. "Image" may appear 30 times in an article about image compression, which is normal and not a problem. A two- or three-word phrase appearing 30 times is usually a real repetition issue.

Start with two-word phrases. If the same pair dominates, look at the surrounding sentences. If nothing stands out, move to three words and see whether a phrase is being repeated verbatim rather than simply discussed.

#Reading the density figures

The tool flags two extremes:

• Above 3.5% — the phrase is a large share of the text. In a short piece this can happen naturally, but in a long article it usually means keyword stuffing or a template that repeats a phrase. Rewrite some occurrences or remove them.
• Under 0.25% with at least two mentions — the phrase is present but thin. Either the page does not really cover it, or the mentions are incidental. Decide whether to build the topic out or drop the phrase from your target list.

Those thresholds are heuristics, not rules. A 300-word product page may legitimately use the product name at 4%, because there is very little else to say. A 3,000-word guide with the same phrase at 4% would be exhausting to read.

#Common problems and fixes

Every heading and link repeats the same phrase. Density is counting navigation and boilerplate, not just prose. Paste the main content only. A page's raw HTML will always look over-optimised because menus and footers repeat.

The phrase I care about is not in the table. It may be filtered out by the minimum word length, split by punctuation, or absent because the tool lowercases and tokenises on whitespace. Check the exact wording in the text. A phrase with a line break inside it will not be counted as one phrase.

Stop-word filtering changed the phrase. With stop words ignored, "how to compress an image" may collapse to "compress image". Turn the filter off if the exact wording matters.

Short words disappeared. A minimum word length of 4 removes png, web, pdf and other short but meaningful terms. Lower it when those terms are the subject.

The counts look low for a long text. Longer texts dilute density. A phrase used 10 times in 5,000 words is only 0.2%. That may be appropriate; density alone cannot tell you whether coverage is good.

The same word appears in different forms. "Compress", "compressed" and "compressing" are separate tokens. The tool does not stem or lemmatise, so it will not merge them. That is deliberate: exact repetition is what reads badly.

The text is over 200,000 characters. The input limit is there to keep the browser responsive. Split long documents into sections and check each one.

#Tips for using density sensibly

• Check after editing, not before. Density is a diagnostic. Write the page for the reader, then look for accidental repetition.
• Judge phrases, not single words. A repeated single word is often unavoidable; a repeated four-word phrase is usually a writing habit.
• Compare against the word count. If a phrase is 2% of a 400-word page, it appears about eight times. Eight mentions in four paragraphs will feel repetitive even if the percentage looks modest.
• Use stop-word filtering to reveal meaning. Ignoring the, and and of leaves the words that carry the topic. Turn it off when you need exact phrase counts.
• Do not rewrite to hit a number. Cutting a useful mention to reach 2.5% is the tail wagging the dog. Rewrite only where the repetition is visible to a reader.
• Pair it with a word count. The word counter gives you the length and reading time; the density checker gives you the repetition. Together they answer "how long is this and what does it keep saying?"

#Where density fits in an on-page review

Density is one small part of on-page work. The title, meta description, headings, canonical tag and structured data matter more, and none of them are visible in a density table. If you are auditing a page, run the meta tag analyzer as well. It checks the title and description lengths, robots directives, Open Graph tags and H1 count from pasted source.

When you are naming the page, the slug generator turns the title into a clean URL without repeating the same words in a clumsy form. Keeping the slug short and the body natural is a better use of time than tuning density.

#Frequently asked questions

#Bottom line

Keyword density is a mirror, not a mandate. Use it to catch repetition you stopped noticing, and to confirm that a page actually discusses the thing it claims to be about. Start with the keyword density checker, then use the word counter for length and the meta tag analyzer for the tags a reader never sees.