A single content score is close to useless. A site scoring 6/10 might be deep but badly structured, or beautifully organised and completely generic. Those need opposite remedies, and one number cannot tell them apart.

The alternative is to score independent dimensions and read the shape rather than the average. Below are the eight we use, what each measures, and what a low score on each one actually means.

1. Depth

Does the page answer the question completely?

Depth is not word count. A 600-word page that fully answers a narrow question is deeper than a 2,500-word page that circles a broad one.

The test is the stopping point: find the sentence where the page stops explaining and starts summarising. On thin content that happens early — typically right after the first subheading, where the hard part would have started.

A low score means: the page was written to exist, not to answer.

2. Topical authority

Does the site cover this subject as a body of work?

One good page on a topic surrounded by nothing ranks worse than one good page surrounded by twelve related ones, because coverage is itself a signal of expertise. See how to tell whether you have it.

A low score means: you have posts, not clusters. The fix is structural, not editorial.

3. Search-intent fit

Does the format match what the searcher wanted?

Someone searching “X template” wants a template. Serving them a 2,000-word essay about the value of templates fails, regardless of essay quality. See matching pages to what people meant.

A low score means: the page may be excellent and still never rank, because it is answering a different question.

4. Differentiation

Could a competitor have published this?

The substitution test: swap your brand name for a competitor's throughout. If the page still reads as entirely plausible, it contains nothing only you could say.

This dimension has become the most important of the eight, for a structural reason. Competent generic content is now essentially free to produce, so its supply is unlimited and its value approaches zero. What remains scarce is content containing something the writer had to know.

A low score means: you are competing on volume in a market where volume is free.

5. Credibility

Are the claims checkable?

Specific, attributed, dated claims. “Studies show” is not a citation. Numbers without sources read as invented, and often are.

Credibility also covers whether the page admits limits. A page claiming a method works universally is less credible than one naming where it does not.

A low score means: readers who know the subject will not trust the rest of the page.

6. Structure

Can a reader find the answer without reading everything?

Descriptive headings, scannable lists where the content is a list, and a hierarchy that reflects the actual argument rather than decorating it.

A structural tell: headings that say nothing (“Introduction”, “Key considerations”, “Conclusion”) usually sit above paragraphs that also say nothing.

A low score means: the content may be good and unreadable, which functionally is the same as not good.

7. Freshness

Is it still true?

This is topic-dependent and often mishandled in both directions. A page on a stable framework does not decay; a page on pricing or tooling decays within a year. Slapping a current date on unchanged content is a worse failure than an old date honestly displayed.

A low score means: the page is actively misinforming readers, which costs more than absence would.

8. Conversion path

Does the page lead anywhere?

An educational page that ends flat has done work for someone else's funnel. The path should be relevant — an article on battlecards linking to a battlecard product, not to a generic homepage.

A low score means: your traffic is real and monetised by nobody.

Reading the shape

Scores are unweighted by default: each dimension counts equally until you have evidence for a different weighting on your site. Weighting invented up front usually encodes the auditor's preferences rather than the site's reality.

What matters is the profile, and a few common shapes have standard diagnoses:

  • High depth, low differentiation. Thorough coverage of what everyone else already covers. Fix by adding original material, not more pages.
  • High differentiation, low structure. You know things nobody else does and nobody can find them. The cheapest fix on this list.
  • High everything, low intent fit. Excellent content aimed at queries nobody makes. Re-target rather than rewrite.
  • Flat mid-range across all eight. The most common and most dangerous profile: nothing is wrong, nothing is compelling, and there is no obvious thing to fix.

That last shape is why the eight scores matter more than their average. An average of 6 tells you nothing. Eight sixes and a two tell you exactly where to spend the quarter.

Our SEO & Content Audit scores a site against these eight dimensions and returns the profile plus a 90-day plan built from the weakest two.