Skip to content
search-engine-optimizedby Netsolution
  1. Home
  2. Page analysis
  3. Keyword Density Checker

Keyword Density Checker

See what your page actually talks about, rather than what you thought it talked about.

Analysed locally on the text extracted from the page. No text is kept after the calculation. No usage limit.

What this tool is for

Let us start with what this tool is not for: hitting a percentage. Keyword density is not a ranking factor, and the thresholds circulating for twenty years have never rested on anything.

What it does reveal is useful. It shows the gap between the subject you thought you were covering and the one you actually wrote. You believe you have written a page about emergency plumbing, and the ranking brings up “quote”, “price” and “callout” — the central word appears three times. Google reads the text as it is.

It also detects the opposite fault: the term hammered forty times in a six-hundred-word text. That is no longer a useful page, it is a manipulation signal, and above all it is unreadable for a human.

To remember: this tool is a revealer, not a target. You use it to diagnose a text, never to calibrate one.

How to read the results

The ranking of isolated words gives the dominant lexical field. The top five terms should match the subject of the page. If they do not, the text is about something other than what the title announces, and Google will rank it accordingly.

The two- and three-word phrases often say more. They reveal the actual formulations, the ones people type. A phrase absent from the list while it is central to your positioning signals a gap to fix.

The percentage reads as an order of magnitude, not to the decimal. Below 0.5 %, the term is marginal in the text. Between 0.5 and 2 %, the frequency is natural. Past 4 %, the repetition becomes noticeable when reading, which is the real warning sign — far more than the figure itself.

The distribution counts as much as the frequency. A term concentrated in the first paragraph then absent from the rest suggests an optimised introduction laid on content about something else.

Common mistakes

Aiming at a percentage. That is the fundamental misreading. Adding a word to go from 1.8 to 2.2 % produces no effect, other than a clumsier sentence. Write to be understood, measure afterwards.

Analysing a whole page, menus included. A forty-link menu repeated everywhere inflates terms that have nothing to do with the content. Always check what was included in the calculation.

Ignoring synonyms. Google has associated related terms for a long time. A text alternating “car”, “vehicle” and “motor” covers the subject better than one repeating the same word — even if its density is lower.

Comparing your density with a well-ranked competitor’s. Correlation is not cause. They are first for reasons that have nothing to do with this measurement: their authority, their age, their links. Copying their density does not copy their ranking.

Confusing density with coverage of the subject. What most pages lack is not repetition of the main word, it is the absence of the related questions.

Going further

Frequently asked questions

What density should I aim for?

None. There is no optimal threshold, and Google does not use this measurement as a factor. The figures in circulation — 2 %, 3 % — are forum inventions from the 2000s. Write naturally: if the subject is covered, the word will appear on its own at a coherent frequency, generally between 0.5 and 2 %.

What is this tool for, then?

Detecting two anomalies, not hitting a target. An abnormally high density reveals a text mechanically repeating a term, which reads badly and can be perceived as manipulation. A zero density on a word you thought you were covering reveals a gap between your intention and what you actually wrote.

What are two- and three-word phrases?

They are the most frequent groups of consecutive words. They are often more instructive than isolated words: “google ranking” says far more than “google” and “ranking” counted separately. It is also where writing tics show up, those formulas that come back without you noticing.

Are stop words counted?

They are excluded by default. Articles, prepositions and auxiliaries — the, of, for, be — make up close to half a text and would crush everything else. You can put them back if you are analysing a writing style rather than a lexical field.

Should menu and footer words be counted?

No, and it is a classic source of error. A menu repeated across three hundred pages artificially inflates certain terms and completely distorts the reading. The tool isolates the main content when the page structure allows it, and tells you what was excluded.

Does Google understand synonyms?

Yes, and has for a long time. Google associates “car” with “automobile”, “doctor” with “physician”, and understands a text’s subject without requiring the exact term to be repeated. That is precisely why chasing a density has become pointless: better to cover the subject’s lexical field than hammer one word.