Keyword density is the percentage of words in a page that are your target keyword. Divide the number of occurrences by the total word count and multiply by a hundred. A 1,000-word article mentioning "image compressor" twelve times has a density of 1.2%.
For roughly a decade this number was treated as a dial you could turn to rank. It never really was, and today it certainly is not. But dismissing it entirely is a mistake too — as an editing check it catches two genuine problems that still cost pages traffic.
What changed
Early search engines matched strings. More occurrences of a term meant a stronger topical signal, so people gamed it: white text on white backgrounds, footer keyword lists, paragraphs that read like a thesaurus having a breakdown.
Modern systems process language, not strings. They resolve synonyms, understand entities, infer intent from context and evaluate whether a page actually answers the query. Google's spam policies explicitly name keyword stuffing as a violation, and their guidance for years has been to write for people.
So density is no longer a lever. What it remains is a diagnostic — a fast, numeric way to spot when a draft has drifted.
The two problems it catches
Dilution. You set out to write about "image compression" and ended up with an article where the phrase appears twice, buried in paragraph nine, while the top terms are "website", "photos" and "loading". A reader would not confidently tell you what the page is for, and neither will a crawler. If your target term is not in the top ten most frequent meaningful words, the page is unfocused.
Over-optimisation. The opposite failure, usually from writing to a density target. Sentences bend awkwardly around an exact-match phrase, the same six words repeat in every subheading, and the prose reads like it was assembled rather than written. Readers bounce, and that is the signal that actually matters.
What a healthy range looks like
There is no official number, and anyone quoting one as fact is guessing. From working on a lot of content, the observed pattern for pages that rank well:
- Primary keyword: 0.5%–2.5%. Roughly five to twenty-five occurrences per thousand words.
- Above 3%: re-read it out loud. It will almost certainly sound repetitive.
- Below 0.3%: the topic may not be clear enough.
Shorter pages naturally run higher — a 300-word product description mentioning the product four times sits at 1.3% and reads perfectly normally. Longer pages run lower. Treat the range as a sanity check, not a target.
Placement beats frequency
Where a term appears matters far more than how often:
- Title tag, near the front.
- H1, once, naturally phrased.
- First 100 words — establishes topic for readers and crawlers immediately.
- One or two subheadings, where genuinely relevant.
- Image alt text, describing the image honestly.
- URL slug, short and clean. The URL Slug Generator handles this.
- Meta description — not a ranking factor, but bolded in the SERP, which lifts click-through.
Hit those and the body density looks after itself.
Write for the topic, not the phrase
Search engines evaluate topical coverage, so the strongest pages use a natural spread of related vocabulary rather than one repeated string. An article about image compression should also naturally contain: file size, JPEG, WebP, quality settings, lossy, lossless, page speed, Core Web Vitals, resizing, metadata.
This vocabulary is not a trick — it is simply what comprehensive coverage produces. If you find yourself unable to introduce related terms, the page probably lacks depth rather than keywords.
Practical way to find them: search your target query, read the top three results properly, and note the concepts they cover that you have not. Also mine the "People also ask" box, which is a direct readout of related intent.
How to audit a draft in five minutes
- Paste the draft into the Keyword Density Checker.
- Enable stop-word filtering so "the" and "and" do not dominate the table.
- Look at the top ten terms. Do they describe your topic? If a stranger read only that list, would they guess the subject? That is the whole test.
- Check your primary keyword's position and percentage.
- Read the two-word and three-word phrases — these often reveal whether you are covering subtopics or circling one idea.
- Fix what the data shows, then read the page aloud. Anything that sounds repetitive out loud is repetitive.
The checker runs entirely in your browser, so unpublished drafts and client work stay private.
Do not optimise the wrong thing
Density is a low-order concern. If a page is not ranking, these matter far more:
- Search intent mismatch. If the top results are comparison tables and you wrote an essay, no amount of density fixes it.
- Thin coverage. Answering less completely than everything else on page one.
- No unique value. Original data, real examples, screenshots, a working tool — something the other results do not have.
- Technical problems. Slow pages, no mobile usability, broken canonicals, blocked crawling.
- No internal links. Orphaned pages struggle. Link to new content from relevant existing pages.
- No external authority. Some queries simply require links.
Fix those first. Then use density as the last editorial pass before publishing.
A short workflow
Draft freely without thinking about keywords at all. Then run the Keyword Density Checker, confirm length with the SEO Word Counter, generate your head block with the Meta Tag Generator, and check the listing with SERP Preview.
Write for the reader, verify with the numbers. In that order — never the reverse.