Sector page
Dental
Implants, ortho and cosmetic. Private research, at night, in silence.
AEO for clinical practices
The Clinical Knowledge Programme
Health questions are asked in private, at night, to a machine that won't judge or sell. Whoever's knowledge that machine is built on decides which clinic gets the call in the morning.
Answer engine optimisation for private healthcare providers. This is the programme; the sector pages are how it's applied. Built and run by Simon Young, Doncaster and South Yorkshire, working nationally.
Sector page
Physiotherapy & MSK
Weeks of self-triage before anyone books. That is the whole opportunity.
Sector page
Audiology
Two people in every sale, and the adult child is the one researching.
Last updated 12 August 2026
What changed
Search didn't die. The click did.
Google's AI Overviews now appear on roughly half of all queries. When one appears, click-through on the top result falls by somewhere between a third and two thirds depending on query type. Around six in ten searches now end without anyone visiting a website at all. Rankings hold. Impressions hold. Sessions collapse.
Healthcare feels this more sharply than most sectors, and for a reason worth understanding properly. Almost every clinical purchase begins with a question the person is too self-conscious, too frightened or too embarrassed to ask a human being. Is this serious? Am I being overcharged? Will it hurt? Is it too late? Will they judge me for leaving it this long?
That research phase used to spread itself across a dozen websites, several of which might have been yours. Now it happens inside one conversation, and by the time the person surfaces into the real world they've already decided what good looks like, what it should cost, and which two or three providers are worth ringing. If you weren't in the conversation, you aren't on the list, and no amount of paid search at the bottom of the funnel buys back a decision that's already been made.
And the traffic hasn't moved channel. In sector after sector the clicks stopped rather than migrated: AI referral traffic is still a low single-digit share of all web visits. “We'll pick up our AI traffic instead” is not a plan. The plan is to be inside the answer.
Why clinical is different
A higher bar, and that's good news
Health is a category where models apply a visibly higher bar before they'll repeat anything. They want a named, qualified, verifiable person behind the claim. They want corroborating sources that agree with each other. They want no contradictions between what you say about yourself and what the public record says. Where those things are missing, an assistant gives a careful non-answer and tells the person to consult a professional, without naming one.
That bar is a barrier to entry, and it happens to be one a real practice with real registered clinicians can clear and a content farm cannot. In most consumer sectors, the well-funded competitor with a content budget wins. In clinical, the qualified practice that publishes carefully wins, because qualification is machine-checkable and a budget isn't.
01
It compounds
Every answer you publish and every third-party source that corroborates it makes the next citation easier to win. Traditional SEO decayed the moment you stopped. This accrues.
02
It's defensible
Once a model is confident your practice is the authority on a treatment in your catchment, a competitor can't outbid you, because there's nothing to bid on. They have to build a deeper, better-corroborated knowledge base than yours, from behind, while you keep building.
03
It's empty right now
Ask an assistant to recommend a provider in your discipline in your town. You'll get a corporate group, a national comparison site, whichever clinic has an old directory listing, and a careful disclaimer. In most towns in Britain there is no named local incumbent.
The window
The land-grab period in any new discovery channel runs eighteen months to two years, and this one started roughly a year ago. What makes it unusual is that being early doesn't just help, it entrenches. Because models weight corroborated, long-standing, consistently-agreeing sources, the practice that establishes itself as the answer becomes the reference point later entrants have to dislodge. Being first on Google meant being first until the next algorithm update. Being first here means being the thing the clinic down the road has to argue with.
The clinical architecture
What doesn't change between disciplines
This is most of what actually moves a model, and it's the same whether you fit implants, hearing aids or a rehab programme.
- Named-clinician entities, not anonymous bylines. Every clinical asset is attributed to a named, qualified individual with registration number, qualifications and role marked up in structured data: GDC, RCVS, HCPC, GMC, GPhC and the rest as applicable. Models cite people as readily as brands, and in healthcare the named human carries most of the trust.
- Registration and record consistency. Practice name, address, registration details, regulator rating and clinician list have to say the same thing everywhere a model can read them: your site, the register, review platforms, directories, professional bodies, Companies House. Contradictions are the single most common reason a model stays vague about a provider.
- Review dates and versioning. Every clinical asset carries a named reviewer and a review date, and gets revisited on a schedule. A compliance good habit that happens to be a strong trust signal.
- Clinical sign-off before publication. Nothing about treatment goes live without your named clinician approving it. I write it, they correct it, it publishes with their name on it.
- A content standard that survives the regulator. No guarantees of outcome, no superlatives, no before-and-afters used as a promise, a clear line between general information and individual advice, risk and limitation acknowledged rather than buried.
- Symptom-side and decision-side, both. Every discipline has two question sets: the person who doesn't yet know they need you, and the person choosing between providers. Almost every practice website addresses only the second.
One honest warning
Five questions to ask any supplier
There's a wave of “AEO” out there that's a rebranded SEO retainer with a new sticker. The tells are easy.
- They conflate AEO and GEO and can't tell you the difference.
- They promise multi-LLM citation tracking but can't produce screenshot evidence against a defined prompt set.
- They treat schema as a one-off setup.
- They tell you third-party authority doesn't matter, usually because they can't do it.
- Clinical-specific: they'll publish clinical claims with no named clinician, no review date and no regard for your regulator's advertising rules. That isn't only a compliance risk, it's the fastest way to be ignored by a model that's cautious about health.
Ask me those five questions. Then ask the next agency the same five.
Why me
I've been writing about answer engine optimisation since October 2019, more than three years before ChatGPT existed. I followed it that November with a piece called “The Death of the SEO Industry.” At the time it read as a provocation. It matters because it means I built this thinking from first principles, from how machines decide what to trust and repeat, rather than reverse-engineering it from a blog post in 2024. The dated receipts are published.
What I'm not is a clinician, and I won't pretend otherwise. I research your clinical content, structure it, and put it in front of your named clinician to correct and approve, then I build the architecture that decides whether a machine trusts it. That division of labour is the whole model, and any supplier claiming deep clinical knowledge across six disciplines is telling you something that isn't true.
The three levels
Published prices, per month, excluding VAT
Three-month minimum, then rolling with 30 days' notice. A twelve-month commitment takes 10% off. Onboarding and build is £995 one-off, waived on twelve months.
Level 1, Visibility
£1,250 / month
Right now, when someone in your town asks an AI about your treatment, you don't come up. This fixes that.
- Baseline and question bank: 30 to 40 high-intent prompts across symptom-side and decision-side, tracked monthly across ChatGPT, Google's AI surfaces and Perplexity with screenshot evidence.
- Entity foundation: schema, named-clinician entities with registration verification, record consistency across directories and review platforms.
- 8 knowledge assets per month, answer-shaped, in your voice, clinician-reviewed and dated.
- Google Business Profile discipline: weekly activity, category correctness, review cadence, attribute completeness.
- Monthly report and review call, including who appears in your place where you don't.
Honest expectation: first movement in citation rates typically shows at 30 to 60 days. Meaningful shortlist presence for your flagship treatment by month four to six.
Level 2, Recommendation
£1,995 / month
Being findable isn't the game. Being named is. This gets you into the two or three practices someone actually rings.
- Everything in Visibility, plus 100 to 120 prompts across six engines tracked fortnightly with evidence.
- Up to three treatments across three locations, architected so they don't cannibalise each other.
- 15 knowledge assets a month, including the full symptom-side set.
- Active third-party citation building across professional bodies, sector directories, regional press, review platforms and referral relationships.
- Fortnightly competitor citation analysis, named-clinician authority building, and a fortnightly working session.
Honest expectation: consistent shortlist presence for your primary treatment by month four to six. Measurable displacement of at least one incumbent by month six to nine.
Level 3, Dominate
£3,495 / month
When anyone in your region asks about your treatment, yours is the name that comes back. Every engine. Every time.
- Everything in Recommendation, plus 30 fully researched pieces a month, each filling an identified gap.
- 250+ prompts, weekly monitoring, live dashboard.
- Unlimited treatments and locations in agreed scope.
- An editorial placement programme across regional and professional press, plus your own original data asset published annually under your name.
- Competitor displacement campaigns on named prompts, area exclusivity in your treatment category and catchment, weekly reporting and a quarterly on-site.
Honest expectation: category ownership in a defined treatment and catchment within six to nine months. Results depend most on clinician time for review and comment.
The audit, £495
The front door, and available on its own
- Your current citation position across all major engines on 25 defined prompts, with screenshots.
- What the machines currently say about your practice, including tone, accuracy and anything out of date.
- Which providers are named in your place, and why.
- Your entity, schema and clinician verification gaps, listed and prioritised.
- A twelve-month roadmap: the questions worth owning, in the order worth owning them.
- Delivered in five working days, in plain English, yours to keep whether you hire me or not.
Credited in full against your first month if you proceed within 30 days. Some practices will read the audit and conclude their catchment is too crowded to justify a retainer. I'd rather they found that out for £495 than for £12,000.
Questions clinical practices ask
Straight answers
Isn't this just SEO with a new name?
No, and the difference is measurable. SEO gets you a link on a page a shrinking number of people click. This gets you named inside the answer, which is increasingly the entire interaction. They share some plumbing, they don't share an outcome. We were writing about answer engine optimisation in 2019, it isn't a rebrand.
Will an AI actually recommend a specific healthcare provider?
Increasingly, where there's a well-corroborated, clearly-qualified, consistently-described local entity to name, and increasingly likely to give a vague non-answer where there isn't. The whole job is being the practice it's confident enough to name.
Is this compliant with our regulator's advertising rules?
It's built to be, and the content standard is the mechanism. Your named clinician approves everything before it goes live. If a supplier isn't asking you about this, they'll cost you more than they make you.
We already rank well on Google. Isn't that enough?
It was. Ranking well now means winning a bigger share of a shrinking number of clicks. It still helps here, because it's part of what models read, but on its own it doesn't get you named, and it doesn't survive a patient who never opens a results page at all.
How do I know it's working?
Screenshots. The same defined prompts, every cycle, every engine, with evidence of where you appear and who appears instead of you. If a supplier can't show you that, they aren't measuring, they're describing.
How much does the Clinical Knowledge Programme cost?
Three levels, per month excluding VAT: Visibility £1,250, Recommendation £1,995, Dominate £3,495. Onboarding £995 one-off, waived on twelve months. The front door is the £495 AI Visibility Audit, credited in full against your first month if you start within 30 days.
I turn down more practices than I take on. If your catchment is too crowded for this to pay back, or your clinicians won't give me the review time it needs, I'll tell you and walk away. I'd rather have five clients who own their categories than fifty who move sideways.
Find your discipline
Sector pages
Sector page
Dental
Implants, ortho and cosmetic. Private research, at night, in silence.
Sector page
Physiotherapy & MSK
Weeks of self-triage before anyone books. That is the whole opportunity.
Sector page
Audiology
Two people in every sale, and the adult child is the one researching.
Not listed? Veterinary, medical aesthetics and the rest of regulated clinical practice run on the same architecture. Start with the £495 audit and we'll find out what the machines say about you today.