What keyword density actually measures
Density is the count of a word (or phrase) divided by the total number of words on a page, expressed as a percentage. If "leather backpack" appears 6 times in a 400-word post, density is 6 / 400 = 1.5%.
Density was a strong ranking signal in the 2000s. Today, Google's models understand semantic relevance — the words around your keyword matter more than the frequency of the keyword itself. That said, density is still the fastest sanity check that:
- You have not accidentally omitted the keyword entirely.
- You have not stuffed it to the point where it reads badly.
- The bigrams and trigrams around it match search intent.
Everything runs in your browser. We tokenize, filter stop words, count n-grams, and expose reading and speaking time — all live.
The bracket that still works
| Target density | What it usually means |
|---|---|
| < 0.5% | Keyword under-represented — Google may not know what the page is about. |
| 1% – 3% | Sweet spot for most content. |
| > 3.5% | Reads as spammy. Fine on tag pages, bad on articles. |
Anything outside this bracket needs a human to look at it, not a rule to enforce it.
Bigrams and trigrams — the underrated tab
Single-word density is noise. The magic is in the two- and three-word phrases:
- "leather backpack" — a head term. Good density = high commercial intent.
- "handcrafted leather backpack" — a long-tail phrase. Even a 0.5% density is a strong signal.
- "backpacks built to last" — an outcome phrase. Great for the description meta tag.
Use the Bigrams / Trigrams tabs to spot the unintentional phrases you keep writing. Those are often your strongest keywords.
Reading & speaking time
We assume:
- 225 words / minute for silent reading — the median for adults on the web.
- 150 words / minute for speaking — average English podcast pace.
Use the reading time for content dashboards; use speaking time when repurposing an article into a script.
Common mistakes
- Optimizing for a single keyword. Optimize for the intent cluster — 5–10 related phrases per article.
- Stripping stop words too aggressively. "How to make leather bags" and "leather bag making" have different intents.
- Ignoring the title and H1. Density counts them exactly once — but a keyword in the H1 is worth 50 occurrences in body text.
- Chasing 3%. Read the article out loud. If it sounds like a robot, back off.
- Not tracking a target phrase. Use the input at the bottom to check any exact phrase, single or multi-word.
FAQ
What density does Google prefer? No published target. Between 1% and 3% for a primary phrase is a safe range.
Should I remove stop words? For measuring density, yes — they inflate the denominator. For understanding intent, keep them in bigrams.
Does this tool understand semantic variants? No — it counts literal strings. Use it as a fast sanity check, then read the article yourself.