Skip to content
TextLab

Free keyword density checker

Keyword density checker for phrases, not just words

TextLab's keyword density checker takes your page copy and a target phrase and reports how often that phrase appears, what share of the page it occupies, where it first turns up and whether it makes the opening hundred words. Free, no account, and it works from pasted copy rather than crawling your site. Alongside it sit the top twenty-five single words, pairs and triples — and one straight answer: there is no percentage you are supposed to hit.

  • 100% free
  • No signup
  • 1, 2 and 3-word phrases
  • Target phrase tracking
  • No magic percentage
density for your target phrasetype a phrase below to check it against 0 words

Type the phrase you are actually trying to rank for. It is matched as consecutive words, so “running shoes” counts only where those two words sit side by side.

Density counts the words a phrase occupies, so five uses of a two-word phrase in 1,000 words is 1.00%.

The twenty-five commonest phrases of the length you pick will be listed here, each with the share of the page it takes up and how far in it first appears.

There is no percentage to aim for. No search engine publishes a target density, and none has counted raw repetitions as a ranking signal for well over a decade — the ranking maths saturates, so the tenth use of a word is worth a fraction of the second. Treat the numbers above as a description of how your draft reads, not a dial to turn.

How to check keyword density

The third step is mostly about knowing when to stop optimising.

  1. Paste the copy a visitor reads

    Take the body text of the page — headings and paragraphs — rather than the source. Markup, script and style would all be counted as words and would wreck the percentages, and so would the navigation and footer that appear on every page of your site rather than on this one. If the page is not written yet, paste the draft: this is more useful before publishing than after.

  2. Name the phrase you are actually chasing

    Type your target into the keyword box and the panel beside it fills in: how many times the phrase appears as consecutive words, what share of the page that occupies, how far in it first shows up and whether it lands anywhere in the opening hundred words. Matching is exact and in order, so entering “running shoes” will not count a sentence about shoes for running.

  3. Read the two- and three-word tables, then close the page

    Switch between single words, pairs and triples to see which phrases your copy actually emphasises — the pairs are usually the revealing ones, because that is where real search phrases live. Then act on anything that reads badly out loud and ignore the rest. There is no number in these tables you are supposed to make equal to another number.

Technical specifications

Density formulaOccurrences × words in the phrase ÷ total words in the text, as a percentage — the convention that makes a two-word phrase used 5 times in 1,000 words 1.00%
Phrase lengthsOne, two and three consecutive words, top 25 of each, ranked by count with density and first position
Sentence boundariesPhrases are never counted across a full stop, so the end of one sentence and the start of the next never form a pair
Target phrase panelExact consecutive-word matching, with count, density, the word number of its first appearance and how many times it lands in the opening 100 words
Function-word handlingPhrases beginning or ending on one of 127 common function words are skipped by default, so “of the” and “in the” stay out of the table
Not countedTitle tags, meta descriptions, alt text, markup and navigation — only the body copy you paste in
Guidance givenNone expressed as a target percentage, because no search engine publishes one
PriceFree, no signup, no crawl limit and no page quota

Frequently asked questions

What is a good keyword density?

There isn't one, and any figure you have been given — three per cent, two per cent, one to two per cent — was invented by a tool vendor rather than a search engine. What is true is descriptive: pages written normally about a subject tend to land somewhere between about half a per cent and two and a half per cent for their main phrase, simply because that is what writing naturally produces. Use that as a sanity check on a draft, never as a target to hit, and if a page falls outside it, read the page rather than adjusting the count.

Is keyword density a Google ranking factor?

No. Google has said publicly and repeatedly for well over a decade that there is no ideal density and that repeating a term past the point of natural use does nothing. The underlying maths explains why: relevance scoring saturates term frequency, so the second mention of a word adds real signal, the fifth adds a little, and the twentieth adds essentially nothing. What is still a ranking factor, in the negative direction, is stuffing — and that gets judged by whether the page reads like it was written for a person.

How do you calculate keyword density by hand?

Divide the number of words your phrase occupies by the total words on the page, then multiply by a hundred. The part people get wrong is the numerator: a two-word phrase used five times occupies ten words, not five, so in a 1,000-word page that is 1.0 per cent rather than 0.5. This page uses that convention throughout, which is why a three-word phrase can show a higher density than a single word that appears more often.

What actually counts as keyword stuffing?

Repetition that a reader would notice as unnatural — the same phrase forced into consecutive sentences, lists of city names, or a paragraph at the foot of the page made of variations nobody would say aloud. It is not defined by a percentage, and it never has been; search engines describe it in terms of intent and reader experience. The practical test is to read the paragraph out loud to someone else. If you would be embarrassed, that is stuffing, whatever the number above says.

Should my keyword appear in the first 100 words?

It usually will if the page is genuinely about that subject, which is the point of the counter next to your target phrase. Nobody has to force it: an opening paragraph that answers the question a searcher asked will contain the words they asked it in. Treat a zero in that box as a prompt to check whether your introduction is actually about the topic, not as a slot to fill — and remember the opening also has to work as the answer someone reads in a search result.

Does this include my title tag, headings and image alt text?

Only what you paste. Titles, meta descriptions and alt attributes live outside the body copy and are usually best checked separately, since they have their own length limits and their own conventions. If you want them in the count, paste them in with the body text — but be aware that adding a 60-character title to a 300-word page nudges every percentage on this screen, which is a reminder of how sensitive density is to what you decide to include.

What about synonyms and related phrases?

They are not counted as your keyword, and that is deliberate — this page counts words, not meanings. Modern search engines do handle synonymy and related concepts, so a page about laptops does not need the word laptop wedged in twenty times to be understood as being about laptops. Use the two- and three-word tables to see which related phrases your copy already emphasises; the phrases you find there are often better guides to what the page is really about than the one you typed in the target box.

How density went from ranking factor to writing smell

In the late 1990s keyword density really was the game. The search engines of the day ranked pages substantially on how often a term appeared in them, so the way to rank for a phrase was to use it more than the next site did — hence the white-on-white blocks of repeated text at the bottom of period web pages. What killed the tactic was not a penalty but better arithmetic. Retrieval models had already moved to weighting a term by how rare it is across the whole collection rather than how often it appears in one document, an idea Karen Spärck Jones published in 1972, and by the mid-1990s the Okapi BM25 model added the crucial refinement: term frequency saturates. Under BM25 the second occurrence of a word contributes far less than the first, the tenth barely registers, and there is a hard ceiling no amount of repetition can push past. Once that became the standard, density stopped being a lever. Repeating a phrase twenty times cannot make a page more relevant than repeating it five times, because the formula refuses to count it.

The tactic then became actively harmful. Google's 2011 and 2012 algorithm updates explicitly targeted thin, over-optimised pages, keyword stuffing is named in the search spam policies to this day, and the guidance from Google's own engineers has been consistent for years: there is no ideal density, and the point at which repetition stops helping is roughly the point at which it stops sounding natural. Meanwhile ranking moved on to things density cannot see — whether the page answers the question, how the surrounding site is regarded, what other pages say about it, and increasingly whether a passage can be lifted out and quoted as an answer.

So why keep a density checker at all? Because it is a good instrument for a bad symptom. A phrase at five per cent of a page almost always means sentences bent around a keyword rather than written to be read, and that is worth catching before you publish. The two-word table is the genuinely useful one: it shows what your copy emphasises, which is often not what you intended, and it surfaces the phrases readers might actually search for. Use it as an editing check, then go and fix the writing. If you want the plain counts without any of the search framing, the word frequency counter gives you every word ranked and exportable; if you suspect the page is repetitive rather than keyword-heavy, the unique word counter measures how much of it is spent on words you have already used; and if the real problem is that nobody can get through the sentences, the readability score will say so faster than any density figure.

Unpublished copy is not uploaded to be counted

Every percentage on this page is worked out inside your browser, so a client's draft landing page or an embargoed launch announcement can be checked without being transmitted to anyone — including to us. Nothing is crawled, nothing is logged against a domain, and the target phrase you type is not stored, which matters more here than on most pages: the keyword you are chasing is often the most commercially sensitive thing on the screen.