Keyword Placement and Density Without Keyword Stuffing
How to place a focus keyword in title, H1, meta, and body without stuffing: AGNURA’s density bands, placement flags, auto-detect vs override, and why intent beats percentage targets.
You will be able to place one focus keyword where it helps clarity, keep density in a sane diagnostic band, and avoid keyword stuffing.
Density is diagnostic. Intent and complete, useful content come first. AGNURA AI Assistant scores a primary phrase against title, H1, meta, URL, intro, early body, and image alts. Those rules are heuristics, not Google’s ranking formula.
Hub: What Is an On-Page SEO Audit?. Related: Title Tags and Meta Descriptions, H1 and Heading Structure.
When this applies
- You have one clear phrase for this URL (or AGNURA auto-detected one)
- An audit flagged missing placement, sparse density, or stuffing
- You want natural language, not a percentage trophy
Prerequisites
- Live page open in Chrome
- A real primary phrase the page can honestly support (or leave blank for auto-detect)
- Optional: AGNURA (Run and Export an AGNURA SEO Analysis)
Direct answer
- Choose one focus phrase that matches the page intent.
- Use it naturally in title and H1 when it fits.
- Mention it early in the body and once in the meta when natural.
- Write the rest for humans; do not repeat the phrase to hit a number.
- Run AGNURA; treat density and placement flags as diagnostics.
- Fix stuffing first if density is high; fix missing title/H1 next; leave Low polish for last.
How AGNURA picks and scores the keyword
| Input | Behavior |
|---|---|
| Focus keyword field blank | Auto-detect from meta keywords, title, and H1 suggestions; first suggestion becomes primary (Auto) |
| You type a phrase | Your override wins (Custom) |
| No usable phrase | Keyword density/placement rules are skipped |
Density = phrase hits in main text ÷ word count (approximate), as a percent.
Density bands (current product)
| Condition | Issue / pass | Severity | Guidance in product |
|---|---|---|---|
| 0 hits in body | keyword-missing | Medium | Include the phrase naturally in main content |
| > 3% | keyword-stuffing | Medium | Reduce repetition; aim roughly 0.5–2.5% |
| > 0% and < 0.3% and page > ~200 words | keyword-sparse | Low | Use the phrase a few more times where it fits |
| Otherwise (when phrase is present and not extreme) | keyword-density-ok | n/a | Density looks reasonable |
Docs also describe a rough ~0.3–3% diagnostic window. The stuffing message’s 0.5–2.5% target is coaching, not a second hard gate. Do not treat either band as a ranking rule.
Placement checks
| Location | Missing issue | Severity | Pass id (when found) |
|---|---|---|---|
| Title | keyword-not-in-title | Medium | keyword-in-title |
| H1 | keyword-not-in-h1 | Medium | keyword-in-h1 |
| Meta description | keyword-not-in-meta | Low | keyword-in-meta |
| First ~100 words (and phrase appears somewhere) | keyword-not-early | Low | keyword-early-ok |
| Introduction window (first ~10% of words, min 50 / max 100; page > ~80 words) | keyword-not-in-intro | Medium | keyword-intro-ok |
| Image alt (if the page has images) | keyword-not-in-alt | Low | keyword-alt-ok |
| URL path | (pass only) | n/a | keyword-in-url when path-like match |
Matching is simple substring style on lowercased text (and URL path with spaces as - or removed). Multi-word phrases must appear as that phrase.
Step-by-step
1. Lock the page intent, then the phrase
Write one sentence: “This URL helps people who want ___.”
Pick a focus phrase that a real searcher might use and that the page delivers. If the page cannot support it, change the page or the phrase. Do not stuff a mismatched term.
Expected result: One honest primary phrase.
2. Decide auto-detect vs override
- Blank field: useful for a quick pass; confirm the Auto phrase is the one you care about.
- Type override: use when Auto picks a brand shard or the wrong H1 fragment.
Click chips or re-run after changing the field when you want a different focus.
Expected result: Report shows Auto or Custom with the phrase you meant.
3. Place for clarity (not for coverage bingo)
Priority order that matches severity:
- Title near the start when natural (title guidance)
- H1 as the visible topic (heading guidance)
- Opening section so readers know the topic early
- Meta once, as part of a real pitch
- One relevant image alt if you have images and it describes the image
- URL when you control the slug; do not force ugly URLs for a Low pass
Expected result: A human can see the topic without noticing “SEO keywords.”
4. Write the body for completeness
Answer the query. Use synonyms and related terms. Mention the focus phrase where a teacher would say the topic name. Stop repeating it in every sentence.
Expected result: Useful page first; density second.
5. Read density as a warning light
| Reading | Action |
|---|---|
| Stuffing (> 3%) | Cut repeated phrases; rewrite lists of keyword variants |
| Sparse (very low on a long page) | Add one or two natural mentions in new sections, not a spam paragraph |
| OK band | Leave it; improve content quality instead |
| Missing in body | Add a clear topic sentence; do not dump the phrase 20 times |
Expected result: No Medium stuffing or missing-body flag, unless you chose a bad Auto phrase (fix the phrase).
6. Re-run and stop chasing Low chips
Re-run SEO Analysis on the live URL. Clear Medium placement/density issues that are fair. Accept remaining Low flags when the prose already reads well.
Common failures and fixes
| Problem | Likely cause | Fix |
|---|---|---|
keyword-stuffing | Same phrase in every H2 and sentence | Rewrite; keep 1–2 natural mentions per section max |
keyword-missing | Wrong Auto phrase, or topic only in images | Override phrase or add real body text |
keyword-not-in-title / h1 | Brand-only title/H1 | Add the topic; keep it readable |
keyword-sparse | Long page, phrase used once | Add natural mentions in new sections |
keyword-not-in-alt | Decorative alts only | Describe one real image; do not stuff every alt |
| Phrase in title but flag remains | Slight wording mismatch (tools vs tool) | Use the exact focus string or change the focus |
| Density looks fine, page is thin | Word count / intent gap | Expand useful content; density will not save a hollow page |
Verification checklist
- Intent sentence written before the phrase
- Primary phrase is Auto-confirmed or Custom-set on purpose
- Title and H1 use the phrase when natural
- Opening section introduces the topic
- Body answers the query without repetitive stuffing
- Density not above ~3%; not artificially padded to hit 2%
- Meta and alts updated only when honest
- Audit re-run on the live page
When to escalate
- Product/legal names that must stay exact: do not warp them for density
- Multilingual pages: substring English-style matching may misfire; judge by native editors
- Sitewide title templates blocking the phrase: fix the template system
How AGNURA helps this job
| Step | Without AGNURA | With AGNURA |
|---|---|---|
| Pick a test phrase | Manual guess | Auto from meta keywords / title / H1, or your override |
| Measure density | Hand count | Percent + occurrence count |
| Check placement | Eye scan | Title, H1, meta, intro, early body, URL, alts |
| Spot stuffing | Subjective | Medium keyword-stuffing above 3% |
Compared with common alternatives
| Approach | Fit |
|---|---|
| Stuff to a “2% rule” | Often harms trust and clarity |
| Ignore keywords entirely | Fine for some brands; weaker diagnostic signal |
| Rank trackers as the placement test | Slow; many confounders |
| AGNURA on the active tab | Fast diagnostics on one URL; not a ranking guarantee |
Limits (product-true)
- Density and placement are diagnostics, not ranking proof
- Exact phrase matching (not smart stemming)
- Auto phrase can be wrong; always confirm
- Main-text extraction and SPA timing affect counts
- Does not use Search Console or rank data
- SEO Analysis does not use managed AI for these checks
- Out of scope: backlinks, CWV, full-site keyword maps
Related guides
- What Is an On-Page SEO Audit?
- On-Page SEO Checklist for One Page
- Title Tags and Meta Descriptions: Practical Length and Intent Checks
- H1 and Heading Structure: What an Audit Should Flag
- Readability Scores for Web Content: Useful Guardrail, Not a Ranking Factor
- How to Read an SEO Score Without Chasing 100
- Run and Export an AGNURA SEO Analysis
Soft next step
Open one URL, leave the focus field blank once and note the Auto phrase, then override with your real primary phrase and compare Medium flags before you edit. Install AGNURA if you want that on the tab:
Try AGNURA in Chrome
Draft in Gmail, LinkedIn, WhatsApp, and Reddit. You review every message before send.
Add to Chrome