Keyword density is one of the oldest ideas in SEO and one of the most misunderstood. It is a simple percentage — how many of your words are your target keyword — but writers treat it as a target to hit, a superstition to fear, or a stick to beat their copy with. It is none of those. It is a diagnostic, useful in a narrow band and actively harmful beyond it. Here is what the number actually measures, the range that is genuinely healthy, and how to check yours in about ten seconds.
What keyword density means (the formula)
Keyword density is the percentage of a text's total words that are your target keyword:
(keyword occurrences ÷ total words) × 100
A worked example: you mention your target term 10 times in an 800-word article. That is 10 ÷ 800 = 0.0125, times 100 = 1.25%. Flip the arithmetic around and it becomes a writing budget: if you are targeting 1.5% in an 800-word draft, your ceiling is 12 mentions (800 × 0.015). For multi-word phrases, count occurrences of the full phrase, not individual words — "email subject line" appearing 12 times in 800 words is 12 ÷ 800 = 1.5%, even though it puts 36 keyword words on the page.
The healthy range: roughly 1–2%
The band WordlyCount's guidance uses — and the one most seasoned editors converge on — is about 1–2% density for a primary keyword. Below that, the page may not clearly signal its topic; above roughly 3–4%, text starts reading as keyword stuffing, and stuffing is the classic signal of over-optimization that search engines have spent two decades learning to spot. The verdicts, in table form:
| Density | Verdict |
|---|---|
| Under 0.5% | Under-signaled — you may have written about the topic without naming it clearly enough |
| 1–2% | Healthy — typical of naturally written, focused copy |
| 3–4% | Edge of stuffing — reads repetitive; edit down |
| 5% and up | Stuffing — rewrite for humans |
Notice what the table is really saying: there is a wide safe valley and steep walls on both sides. You do not need to land on 1.37%. You need to stay off the walls.
What stuffing actually looks like
Stuffing is a readability crime before it is an SEO one. It looks like this: "Our word counter tool is the best word counter because our word counter counts words faster than any other word counter." It also looks like paragraph-ending lists of city names, alt text that has nothing to do with the image, and a footer that repeats forty service variants in 8-point type. The test that never fails: read the paragraph aloud. If you can hear the keyword before you finish the sentence, everyone else can hear it too — including the systems that decide rankings.
Density is a floor, not a ceiling
The modern view is that density matters less than writers fear, because modern search is semantic. Search engines now resolve entities, synonyms, and related terms — they understand that a page about "running shoes" that also says "trainers", "sneakers", and "cushioning" is about running shoes without the exact phrase being hammered. What they reward is intent fit: a page that genuinely answers the query the keyword represents, structured the way searchers for that query expect.
So use density as a floor, not a ceiling: it is the check that you have actually named your topic often enough for both readers and machines to be sure what the page is about. If your target term appears twice in a 2,000-word draft, the fix is not to jam in eighteen more mentions — it is to check whether the draft really is about what you think it is. Natural usage beats a number every time, and no ranking is worth sacrificing clarity to raise density by half a point.
How to check your density in WordlyCount
The check takes longer to explain than to do:
- Paste your draft into the WordlyCount counter.
- Open the top-keyword panel — it lists your most frequent terms with their occurrence counts.
- Find your target term's count and divide by your total words, times 100: 10 mentions in 800 words is 1.25%.
- Sanity-check the prose around each mention — if any single paragraph leans on the term twice, that paragraph is where the draft starts sounding synthetic.
While you are there, glance at readability: a Flesch score of 60–70 is plain English, the register most search-facing copy should sit in — we cover that scale in what Flesch-Kincaid readability means. And if trimming repetitions drags your draft off-topic, the deeper fix is usually length itself: a tighter piece says its keyword naturally fewer times, which is one reason the targets in how many words a blog post needs feed directly into healthy density.
When keyword research — not density — is the real lever
Here is the honest ranking of effort. Tuning density from 1.1% to 1.6% is maintenance; choosing a different keyword is strategy. Targeting a term nobody searches for at a perfect 1.5% earns nothing; targeting a term with real volume, achievable competition, and a page that matches its intent earns plenty at 0.9%. Before you spend an afternoon deleting mentions, spend it verifying that the phrase is the right one to rank for — the density dial is the last 5% of the job, not the first.