Keyword Density Checker

How the Keyword Density Checker works

Keyword density checker: a diagnostic, not a target

Keyword density is a diagnostic, not a target. It tells you what a piece of writing repeats, which is genuinely useful — and it has been misread for two decades as a number to optimise.

Density is not a ranking factor

There is no threshold that helps and no percentage to aim for. Search engines moved to statistical language models long ago and have penalised keyword-stuffed text since the Panda update in 2011. Anyone quoting a target of "1 to 2 percent" is repeating advice that was obsolete before most of today's web was written.

Content keyword analysis: what the table is actually good for

Three things. First, catching accidental repetition — a word used forty times in a thousand is a tic, and reading it aloud confirms it. Second, checking that the subject you think you wrote about is the subject the text repeats; if a page about invoice templates never says "invoice" outside the heading, that is worth knowing. Third, spotting vocabulary you have missed entirely: the phrases a reader would search for that appear nowhere.

N-gram frequency counter: phrases beat single words

A single-word count of a technical article is mostly noise, because the words that carry meaning are compounds. Two- and three-word phrases surface the actual subject matter — "reading time", "style guide", "zero-width space" — and they are far closer to how people phrase a search. Start at two words and drop to one only when you want the raw vocabulary profile.

Stop words dominate everything

In English, the, of, and, to and a make up roughly a quarter of any text. Leaving them in produces a table that tells you nothing about the document and everything about the language, which is why they are filtered out by default. For phrase counting, a phrase is dropped when any of its words is a stop word — that removes "of the" and "in a" but also legitimate phrases like "out of stock", so turn the filter off if you are hunting for something specific.

How the density percentage is calculated

Density is the count divided by the number of positions a phrase of that length could occupy. For single words that is simply the word count; for three-word phrases it is two fewer. The denominators therefore differ between phrase lengths, and comparing a one-word density against a three-word one is meaningless.

Word frequency analyser: counting rules

Everything is lowercased before counting, so Invoice and invoice are the same entry. A word is a run of letters and digits, with apostrophes kept inside a word, so don't counts once. Phrases are counted across all positions including overlaps, and they are not stopped by punctuation — a phrase can span a full stop, which occasionally produces an odd-looking entry in the table.

A worked example: the same text at three phrase lengths

A 1,000-word article about invoice templates, counted with stop words excluded.

"invoice" at 2.4% is fine for a page about invoices; the useful finding is that "payment terms" appears six times while "due date", the phrase readers search for, appears nowhere. The three-word density looks tiny only because the denominator is almost the whole document.

Related tools