Content quality scoring compares a page against what currently ranks for its target search and reports where it falls short: what the question needs answering, whether the answer is findable, whether claims are evidenced, and whether the expertise behind it is visible.
Built in-house. It is my own tooling rather than a reseller’s dashboard, which is why it checks the things I act on rather than everything a generic tool can measure.
What it does
The scoring gives every page a verdict rather than a number on its own. Keep it, rewrite it, merge it into something else, or remove it. That verdict is the useful output, because a client with 300 blog posts needs to know which forty to work on, not a spreadsheet of scores.
Underneath it, the checks are about whether the page earns its place for a real search.
What it looks at
| Area | What it looks at |
|---|---|
| Coverage | What the ranking pages answer that this one doesn’t, and what it covers that nobody searches for |
| Answer quality | Whether the question is answered near the top, in a passage that stands on its own |
| Specificity | Real numbers, dates, named features and worked examples, against vague general advice |
| Evidence | Whether factual claims are sourced, and whether the source is primary |
| Experience | Whether the page shows first-hand knowledge or just summarises other pages |
| Named expertise | A real author with real credentials, rather than an unattributed post |
| Duplication | Whether another page on the same site covers the same ground |
| Performance | Impressions, clicks, position and trend from Search Console |
Why E-E-A-T sits inside this rather than beside it
Experience, expertise, authoritativeness and trust aren’t a separate tick-box exercise. They show up as concrete things on a page: whether there’s a named author with a reason to be believed, whether the writing shows someone has actually done the thing, whether claims carry sources, and whether the business behind the page is clearly identified.
Scoring those alongside coverage and answer quality keeps them practical. An unattributed page full of general advice scores badly on both, and the fix is the same piece of work.
Duplication is usually the bigger finding
On most content-heavy sites, the largest single problem isn’t quality, it’s two or three pages competing for the same search. Each is reasonable, none is strong, and the ranking signals split between them.
That’s why the scoring looks across the set rather than page by page. A merge recommendation is often worth more than three rewrites.
Example output
PAGE: /insights/how-long-does-a-claim-take/ VERDICT: Merge. WHY: Two other posts cover the same question. This one has the most impressions and the weakest answer. The other two have the detail. ACTION: Combine into this URL, redirect the other two, keep the timeline table from the newest one.
What I review
Every verdict is checked against the live page and the live search result. A tool can tell you a page is thin, but only a look at what ranks tells you whether thin is actually the problem.
I also check the content tracker before calling anything a gap. On a client running a content programme, a missing page is often already written and waiting for approval, and reporting it as a gap wastes everyone’s time.
Limitations
- A score is a prompt, not a verdict on the writing. Some excellent pages score badly because they target a search nobody makes.
- Competitive context shifts. The comparison is against what ranks today.
- Search Console caps its reports, so performance figures on large sites are a floor rather than a total.
Services it powers
Content quality scoring powers content quality and E-E-A-T, content refreshes and the content side of SEO. See the rest of the tools on the tech stack page.
FAQ
What does content quality scoring measure?
It measures whether a page earns its place for a real search: how fully it answers the question compared with what ranks, whether the answer is findable near the top, whether claims carry sources, and whether the expertise behind it is visible to a reader.
Is E-E-A-T something you can actually score?
Not as an abstract quality, but the things it describes are concrete and checkable. A named author with real credentials, evidence of first-hand experience, sourced claims and a clearly identified business are all things a page either has or doesn’t, and all four are fixable.
What happens to pages that score badly?
They get one of four verdicts: keep, rewrite, merge or remove. Merging is the most common recommendation on content-heavy sites, because the usual problem is two or three reasonable pages competing for the same search rather than any single page being genuinely poor.
See the whole toolkit
The tech stack page lists every tool, and the free assessment includes a look at your existing content.