What the Keyword Density Checker does
This checker counts how often each word, two-word phrase and three-word phrase appears in a piece of content, shows each count as a percentage of all words, and - when you give it a target keyword - checks whether that keyword appears in the title, meta description, H1, subheadings, first 100 words, URL and image alt text.
Paste plain text or a page's HTML and everything is worked out in your browser. You can also enter a public URL instead: the a2z.tools server fetches that page's public HTML once, when you press Analyse, and nothing else is sent.
How to use it
- Paste the article text, or the full HTML of the page if you want the title, headings, meta description and alt text checked too. Alternatively, enter the page's public address.
- Type the keyword or phrase you are writing about, if you have one. Leave it empty to see only the most-used words and phrases.
- Press Analyse. Switch between the One word, Two words and Three words tabs to see which phrases the text leans on.
- Read the prominence checks: each says whether the keyword is present in that place. Add it where a reader would expect it, not everywhere.
- Download the CSV to keep the full phrase tables, or copy the short summary into your brief or review notes.
Reading the results
Density is occurrences divided by the total number of words, times 100. A two-word phrase used 5 times in 54 words has a density of 9.26%.
Share of words multiplies the occurrences by the number of words in the phrase, so the same phrase covers 10 of the 54 words, or 18.52%. Some tools report this figure as density; the table shows both so you can compare like with like.
Heavy repetition is flagged when the target's density is above the threshold you set (3% by default). That threshold is an A2Z rule of thumb for spotting text that repeats a phrase awkwardly. No search engine publishes a target or maximum density.
With stop words ignored, single words such as "the" and "and" are left out, as are phrases that start or end with one. Phrases never run across a sentence boundary.
Worked example: a product description that repeats its keyword
Load the product text example with the target "wool socks". The text has 5 sentences and 54 words: 13, 12, 11, 12 and 6.
"Wool socks" appears once in the first sentence, once in the third and three times in the fourth, 5 times in all. Density is 5 / 54 x 100 = 9.26%, and share of words is 5 x 2 / 54 x 100 = 18.52%. Both are well above the 3% threshold, so repetition is marked Heavy.
In the One word tab, "wool" and "socks" each appear 6 times (11.11%) because "the socks" in the second sentence adds one more "socks" and "merino wool wicks" one more "wool". The fix is obvious once you see it: the fourth sentence, "Wool socks, wool socks, wool socks", adds nothing a reader needs.
What keyword density can and cannot tell you
Google's own guidance on helpful content warns against stuffing pages with keywords and describes keyword stuffing as a spam practice. It does not give a number that is too much, and there is no published ideal density for any search engine. A density figure is therefore evidence about the writing, not a ranking score.
It is still useful in two ways. A very high figure, or one phrase dominating the two- and three-word tables, usually means sentences have been written around a phrase instead of for a reader. A zero in the title, H1 or first paragraph can mean the page never says plainly what it is about.
How this differs from a word frequency counter
The word frequency counter lists every word in a text. This page is aimed at one question - how a chosen phrase is used - so it counts two- and three-word phrases, reports density against the whole text, and checks the places on a web page where the phrase matters: title, meta description, headings, URL and alt text.
Limitations: what the result does not prove
- It measures repetition, not relevance or quality. A page can have a textbook density and still fail to answer the reader's question.
- Matching is exact after lower-casing and removing accents: "sock" does not match "socks", and synonyms are not counted.
- When you use a URL, only the HTML the server returns is read. Content added later by JavaScript, text behind a login and pages that block automated requests are not seen, and the page says when a fetch fails rather than showing zeros.
- From a URL, navigation, header, footer and aside elements are removed and the main or article element is used when there is one. A page with unusual markup may include or exclude text you did not expect.
- The word count uses a Unicode-aware word splitter. Scripts written without spaces, such as Chinese or Japanese, are not split into words and give misleading figures.
Privacy: where your data goes
The analysis runs in your browser. If you choose to load a public URL instead of pasting, a2z.tools fetches that address read-only and returns the text to your tab; private and internal addresses are refused and nothing is stored. Session recording and tag-manager scripts are switched off on this page.
Standards and sources
- Google Search Central - Spam policies: keyword stuffing
- Google Search Central - Creating helpful, reliable, people-first content
- Google Search Central - Influencing your title links
Frequently asked questions
How is keyword density calculated?
Divide the number of times the keyword appears by the total number of words and multiply by 100. For a phrase, this page also shows share of words: occurrences times the number of words in the phrase, divided by the total, which some other tools call density.
What is a good keyword density for SEO?
There is no published figure. Search engines do not state an ideal or maximum density, and Google's spam policies describe keyword stuffing without giving a number. Use the figure to find awkward repetition and missing placements, then write for the reader.
Why does my density differ from another keyword tool?
Tools differ in whether they count stop words, whether they include navigation and footer text, how they split words, and whether a phrase counts once or once per word. Check the share of words column and the stop-word option before comparing.
Can I check a competitor's page by URL?
Yes, if the page is public and returns HTML. The server fetches it once and the analysis runs in your browser. If the site blocks the request or needs JavaScript to show its content, the page says the text is unavailable instead of reporting an empty result.
Does the tool count keywords in headings and the title?
When you paste HTML or use a URL, the title, meta description, H1, H2 to H6 headings, URL path and image alt text are checked separately for the target keyword. Plain pasted text has no such structure, so only the first 100 words are checked.
Why are phrases like "the socks" missing from the tables?
With stop words ignored, phrases that start or end with a common word such as "the" or "of" are skipped so the tables show meaningful phrases. Untick Ignore stop words in Options to count everything.
Last reviewed by the A2Z.Tools team against the sources listed above.
