Content Quality Score
How SitePulse measures content quality across readability, depth, structure, engagement, and technical execution.
The Content Quality score measures how well a page's written content serves its audience — is it readable, substantive, well-structured, and engaging? This is the sixth audit category and the only one that evaluates the writing itself rather than technical signals.
Why content quality matters
Search engines and AI models increasingly use content quality signals to differentiate pages covering the same topic. A page with excellent SEO metadata but thin, hard-to-read content will rank and get cited less than a substantive, well-written competitor. Content quality is also the strongest lever for reducing bounce rate and increasing time on page.
Readability (25 pts)
Readability measures whether a typical visitor can understand the content without re-reading.
| Condition | Points |
|---|---|
| Flesch Reading Ease ≥ 60 (standard/plain) | 10 |
| Average sentence length ≤ 20 words | 6 |
| Average paragraph length ≤ 100 words | 5 |
| No sentences longer than 40 words | 4 |
Flesch Reading Ease bands
| Score | Level | Typical audience |
|---|---|---|
| 90–100 | Very easy | 5th grade |
| 80–90 | Easy | 6th grade |
| 70–80 | Fairly easy | 7th grade |
| 60–70 | Standard | 8th–9th grade (target for most web content) |
| 50–60 | Fairly difficult | 10th–12th grade |
| 30–50 | Difficult | College level |
| 0–30 | Very difficult | Professional/academic |
For most marketing, product, and help content, a score of 60–80 is ideal. Technical documentation and legal content may legitimately score lower.
Technical quality (20 pts)
| Condition | Points |
|---|---|
| Content-to-HTML ratio ≥ 25% (not tag-heavy) | 6 |
| No detected duplicate paragraphs (boilerplate blocks) | 6 |
| Low passive voice usage (≤ 15% of sentences) | 4 |
| No common grammar error patterns detected | 4 |
The content-to-HTML ratio measures how much of the page's raw HTML is actual visible text versus markup, scripts, and styles. A low ratio (< 10%) indicates the page is predominantly technical boilerplate with thin readable content.
Content depth (20 pts)
| Condition | Points |
|---|---|
| Word count ≥ 300 words | 6 |
| Word count ≥ 600 words | 4 |
| Named entities present (people, places, organizations, products) | 5 |
| Topic covered beyond surface level (entity diversity metric) | 5 |
Content depth does not reward padding. A 300-word page that is dense with specific facts scores higher than an 800-word page full of filler phrases and vague generalities.
Structure (20 pts)
| Condition | Points |
|---|---|
| At least 2 H2 headings present | 6 |
| Heading hierarchy respected (no level skips) | 5 |
| At least one ordered or unordered list | 5 |
| No single paragraph exceeding 200 words | 4 |
Engagement signals (15 pts)
| Condition | Points |
|---|---|
| At least one call-to-action (button, link with action verb) | 5 |
| At least 3 internal links to related content | 5 |
| Images or video present | 5 |
Brand Voice & Persona scoring
If your project has Personas configured, the Content Quality score includes an additional Brand Voice dimension that scores how well the content matches your target audience's reading level, tone, and vocabulary.
When personas are active, each persona is scored against the page content and the lowest-scoring persona drives a weighted adjustment to the overall Content Quality score. This ensures content works for your most important or hardest-to-reach audience, not just the average reader.
The persona adjustment is shown separately from the base score in the Metrics breakdown so you can see which personas the content is underserving.
How to interpret your score
| Score | Meaning |
|---|---|
| 90+ | Content is well-written, substantial, and structured for both human readers and AI systems |
| 70–89 | Good overall; likely one or two specific dimensions (readability, depth) underperforming |
| 50–69 | Multiple issues — typically thin content combined with poor readability |
| < 50 | Significant content problems — either very thin, very difficult to read, or both |
How to improve
Readability fixes (highest ROI):
- Break long sentences at the conjunction ("and", "but", "which")
- Split long paragraphs after 3–4 sentences
- Replace passive constructions: "was found to be" → "is", "was performed by" → "X did"
Depth fixes:
- Identify the top 3 questions a visitor would ask about the page topic and answer them explicitly
- Add at least one concrete example, case study, or data point
- Name specific products, people, or places rather than using generic terms
Structure fixes:
- Add descriptive H2 headings every 200–300 words
- Convert any run-on explanatory paragraph into a bulleted list
The Rewrites tab generates reading_level_fix and paragraph_split suggestions targeting the specific sentences and paragraphs that are dragging down your score.
Content Quality and GEO overlap
Several Content Quality dimensions (depth, structure, entity density) directly feed into the GEO score as well. Improving content quality almost always improves GEO simultaneously. The GEO score page has the full rubric for the GEO-specific dimensions.