What is my article score, and how is it measured?

This is part of the Clinical Content System subscription. If your clinic isn't subscribed yet, you can see what the Clinical Content System includes.

Sound familiar?

"How do I know the content is actually good?" "What is this score on my pages?" "Is anyone checking what the AI writes before it goes out?" "Why did my existing pages get scored too?" "Why did my scores change in September?" One standard answers all of these, and it runs through the entire platform.

What's happening

Every article the platform writes for you is scored against a set of AI-search checks before it reaches you, using the checks that apply to its page type. Not a spellcheck and not a vibe: specific checks, each based on research into how AI search engines choose the sources they quote.

The same scoring runs everywhere:

It scored every eligible content page of your website during your AI Visibility Assessment, each page individually, 0 to 100 (legal and boilerplate pages are left out).
It scores every article we write for you.
It guards quality on the way through: a piece has to clear a quality bar and its critical checks before it lands in your Studio as ready, and anything that falls short when you come to publish (like an unsupported credential claim, or a missing review) is flagged so you can put it right 📏.
It holds our own articles to a higher bar: a sentence that could make a reader doubt the care lowers the article's score, and it's listed with a suggestion to reword it so it describes what the care involves. Safety advice and "see your GP" lines never count against an article, and this never stops you publishing.

Why it works this way

AI search engines choose sources on what a page actually says and how easy it is to use. So the checks focus on that:

Answering the related questions patients actually ask. This is the strongest driver we have measured. A page that covers the follow-up questions (how long will this take, when should I worry, what does treatment involve) is far more likely to be quoted than one that doesn't.
Clear, well-sized answers. Sections that give a straight answer in a passage of sensible length, so an AI tool can lift it cleanly.
Evidence citations: real, checkable references.
A named, qualified author and reviewer, so patients can see who wrote it and who checked it.
Internal links that connect the page to the rest of its topic on your site.
Readable without special technology: a page AI tools can read even without running its scripts.
Reachable by AI crawlers: a page the AI tools are actually allowed to visit.

The scoring was rebuilt on 21 September 2026, on measured evidence. Checks that counted the same thing twice were merged, checks with no evidence behind them were dropped, and the "related questions" check above was added. Every clinic's scores moved that day for reasons that had nothing to do with their website. If you're comparing scores from either side of that date, that's why they differ.

And the score is only the final check on a much bigger process. Before a word is written, each article goes through a multi-step research pipeline. The finished piece you read is that research distilled, then scored 🔬.

What to do now

See your existing pages' scores in the Page Inventory at https://content.co-kinetic.com/my-clinic/page-inventory: every page, individually scored, with its band and how many things it has to fix. Your assessment report at https://content.co-kinetic.com/my-clinic/ai-visibility shows the overall picture.
Low-scoring pages are not bad news, they are your plan's raw material: they surface as rewrite actions on your plan at https://content.co-kinetic.com/my-clinic/plan, because a page that already exists is often worth more rewritten than a new page is worth created.
Articles that have cleared the quality checks arrive in your Content Studio at https://content.co-kinetic.com/my-clinic/studio for your review before anything is published. The score protects the floor; your clinical judgement is still the final sign-off.

The question behind the question

"Can I see the full list of checks?" No, and that's deliberate. The checks change as the evidence about AI search changes, and the number that applies depends on the page type, so there isn't one fixed list to show you. What you get is the useful end of it: each page's score, its band, how many things it has to fix, and a content plan that turns those into actions in the right order 🙂.

The next question is usually...

What is the Page Inventory?
How is my content plan organised, and how does it decide what goes where?
Why do I need a reviewer at all?
What happens once my content plan unlocks?

Still stuck? This usually means...

A page you rate highly scored low. Design and clinical quality are not what the checks measure. They measure whether the page answers patients' real questions clearly, and whether AI tools can reach and read it.
Your scores changed around 21 September. That's the scoring rebuild described above, not something that happened to your site.
You disagree with a page's score. Ask us in the chat with the page address and we'll explain what's holding it back.
You want to raise a page's score yourself. You usually don't have to: low scorers surface as rewrite actions on your plan, and letting the plan rewrite them is the designed route.