Why isn't my website cited as the expert answer yet, and how do we get there?

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?

"It says we were cited in 0 of 38 expert answers. Why?" "The NHS keeps getting quoted instead of us." "How do we become the source AI uses?" Being cited as the expert answer is the holy grail of AI search. It's also the hardest thing to win. Here's why, and what gets you there.

What the number measures

Your AI Visibility results count two different things.

Most clinics start at or near zero on expert answers. That's normal, and it's not a sign anything is wrong.

Why it's so hard to win

When a patient asks "what's causing the pain under my kneecap?", AI looks for the most trusted, most complete answer it can find anywhere in the world. Right now that's usually the NHS, the Mayo Clinic, the Cleveland Clinic and similar sites. They have thousands of pages, decades of reputation, and millions of other sites linking to them. A single clinic can't match that on volume, and nobody can guarantee you'll take an answer from them.

So the big health sites will always hold many of these answers. That's the honest picture.

Why it's still worth going for

The big sites win the broad questions. They're much weaker on the specific ones. They write one general page on knee pain. They don't write about knee pain in runners who've just increased their mileage, explained by a named physio who sees it every week. That's where an independent clinic can become the best answer.

AI chooses a source it can trust, and it decides that from:

  1. Depth on a subject. One page on a condition is a single mention. A main page for the condition, with supporting articles that each answer one specific patient question, all linked together, shows AI that your clinic knows this subject inside out. That's how your content plan is built.
  2. The patient's exact question, answered plainly. Pages that answer the question directly, near the top, in the words patients use, are the ones AI quotes.
  3. A named, qualified expert behind it. Every page we write carries the author and reviewer, with their credentials. AI can see who wrote it and why to trust them. The big sites often can't match that personal expertise.
  4. Time and consistency. Each page you publish makes the next one count for more. Reputation in AI search builds the way it does with patients: steadily, through being reliably good on your subjects.

What to expect

This takes months, not weeks, and it moves gradually. You'll usually be named in local answers before you're cited in expert answers, and that's the right order: being named is what brings patients in. Expert-answer citations tend to come first on your most specific subjects, the ones where you've built real depth.

Your results show every answer, so you can see exactly which questions you're winning and which sources AI is using instead.

What to do now

  1. Publish the pieces in your plan, in order. The plan builds depth topic by topic, which is what wins citations.
  2. Send each piece for review. A page that shows its reviewing clinician is a stronger source.
  3. Keep your clinician profiles rich. The named expert behind every page is part of what makes AI trust it. See "Do we have to publish the whole clinician profile on our website?"
  4. Look at which conversations you're tested on. Your AI Visibility results show the four patient conversations. Pages that answer those conversations' questions are the ones that move this number first.
  5. Press Validate & earn on each published piece, so we can check it's live and measure it.

The question behind the question

"Is it worth it if the NHS always wins?" Yes. You don't need every expert answer. Every one you win puts your clinic's name in front of a patient at the exact moment they're trying to understand their problem. And the same depth that wins expert answers is what gets you named when they ask who to see.

The next question is usually...