Healthcare SEO for clinics and medical practices: getting found by patients who want an appointment (without copying Bupa's blog)

Healthcare SEO for a medical practice, a private clinic or a multi-specialty centre is the work that puts the practice on page one, in the local pack and in Google's AI answers when a patient searches for a treatment, a specialty or a clinic in their town. It is one of the few sectors where a single practice still reaches page one with its own website: on "private gp london" the first page holds eight results from private clinics and one directory, with no NHS page and no national chain.

And what do patients actually type?

Not a doctor's name.

They type "cardiologist london", "private doctor near me" (5,400 searches a month in the UK), "dermatologist london" (9,900).

Compared with SEO for any other website, three things change:

• the pages that rank are treatment pages and location pages;
• the Google Business Profile carries as much weight as the site;
• what you're allowed to write is governed by professional guidance and advertising rules.

And yet the most common advice is "start a health blog, like the big brands do".

Meanwhile, in the clinics we see, eight services out of ten have no dedicated page with price, duration and clinician.

Here's what you'll find:

• what we measured on UK medical SERPs in September 2026 (DataForSEO, Google UK);
• why the Bupa model can't be copied, and the London Doctors Clinic one can;
• why Keyword Planner misleads you here (8 keywords out of 20 with no data);
• the three pages that rank and how to link them;
• what you can write within GMC and CAP rules;
• how to get cited by the AI Overview and how to measure bookings.

Dentists and psychologists have their own guides: SEO for dentists and SEO for psychologists.

Let's start with two SERPs.

How patients search: two SERPs that behave differently (and what that means for you)

Medical searches split into two types that Google treats differently: "specialty + town" opens with an AI Overview and a row of directories, "private GP + town" opens with the local pack and no AI Overview. We analysed both on 26 September 2026, desktop, Google UK.

"cardiologist london"

• Opens with an AI Overview, followed by a "Find results on" carousel of directories (Top Doctors, Doctify, Cylex, Doklist).
• Local pack of three listings: Dr Francesco Lo Monaco on Harley Street (379 reviews, rating 5.0), London Cardiovascular Clinic (238 reviews, 5.0) and Heart Health at Welbeck (105 reviews, 4.8).
• Nine organic results: only one directory (Top Doctors, fourth); the other eight are cardiologists, clinics and hospitals, from One Heart Clinic in first place to Cleveland Clinic London and Cromwell Hospital.
• Ads: none.

"private gp london"

• No AI Overview.
• Local pack of three listings: Private Medical Clinic in Portman Square and two sites of Private GP Surgery.
• Nine organic results: eight private practices and one directory, Find A Private GP, in eighth place.
• Several of the ranking pages are service pages, not home pages: home visits, membership plans, same-day appointments.
• Ads: none here either.

Search	AI Overview	Local pack	Directories in top 10	Individual providers	Ads
cardiologist london	yes, at the top	yes, 3 listings	1 of 9 (plus a directory carousel)	8 of 9	0
private gp london	no	yes, 3 listings	1 of 9	8 of 9	0

So where should the work go?

Two different places.

• Specialist search ("specialty + town"): the treatment page has to be clear enough for the AI Overview to cite it, and the directories take a whole row above the local pack.
• Practice search ("private GP + town"): the Google Business Profile and the service pages win.

Two different jobs. You need both.

And the directories?

Top Doctors and Doctify appear twice on "cardiologist london": in the carousel and, for Top Doctors, in the organic results too.

The profile on those platforms is worth keeping, because many patients compare consultants there.

So where's the catch?

If the treatment page on your own site has less content than your directory profile, Google ranks the directory. And the patient who was looking for you by name books through someone else's page.

One limitation to state, honestly: these are two SERPs, on one day, on desktop, with UK-wide location. On mobile or from a specific postcode the local pack changes, and the AI Overview can appear on the second search too.

Recommended reading: how to read search intent from a SERP.

Why the Bupa model can't be copied (and the London Doctors Clinic one can)

Bupa gets most of its top traffic from health information articles with no private treatment and no town in them: it's a publisher's strategy, not a practice's. The model you can copy is London Doctors Clinic, which ranks on its clinics and its services.

"Start a health blog, like the big brands do."

Sounds reasonable, right?

The numbers say no.

Bupa.co.uk:

• 115,336 organic keywords in the UK and an estimated 3,282,465 visits a month;
• among its 50 highest-traffic keywords, 32 are informational and none is a private treatment combined with a town;
• the top ones are "bmi calculator" and its variants (1,000,000 searches a month each), "electrolytes" (165,000) and "lower back pain" searches (60,500), all served by the health information hub and the newsroom;
• the place names in its top 50 are care homes and hospital directory pages (John Radcliffe, St George's, Cromwell Hospital), not clinic pages.

That's a publisher's strategy: thousands of articles, a clinical editorial team, years of publishing and a national brand.

And the person in Leeds who reads "what are electrolytes" is not going to book a consultation in London.

The realistic model: London Doctors Clinic

London Doctors Clinic: 2,663 organic keywords and around 49,700 estimated visits a month.

Sixty-six times less than Bupa.

So why should it be the model?

Because of where that traffic comes from. It ranks:

• second for "doctors near me" (49,500 searches a month) and "gp private near me" (14,800, CPC £3.89);
• first for "private general practitioners london" (2,900, CPC £7.41) and "private x-ray near me" (1,900).

In its top 50 keywords, 19 contain "near me" or a place, and the location pages rank first for their own names: "london doctors clinic canary wharf" and "london doctors clinic london bridge" (1,600 searches each), "london doctors clinic kings cross" (1,300).

It has one blog post that brings a lot of traffic ("ibuprofen vs paracetamol"). The rest of the top 50 is service pages, blood test pages and clinic pages.

Domain	Organic keywords	Estimated visits/month	Informational kw out of 50	Kw with "near me" or a place out of 50	Brand kw out of 50
bupa.co.uk	115,336	3,282,465	32	11 (care homes, hospital directory)	7
londondoctorsclinic.co.uk	2,663	49,672	32	19	4

London Doctors Clinic has a page for each clinic and a page for each service, and it ranks for searches that end in an appointment.

It's the model for a network of clinics. And, on a smaller scale, for a practice with a single site.

Recommended reading: how company size changes SEO strategy.

Which keywords to target (and why Keyword Planner misleads you here)

In healthcare Google Ads returns no volume or CPC for many medical keywords: of the 20 UK treatment, specialty and practice keywords we checked, 8 came back empty. The demand exists anyway.

The 20 keywords were things like "cardiologist london", "private gp london", "private blood test london", "dermatologist near me".

12 have volume and CPC. 8 come back empty, including "private cardiologist london", "private gp london", "private gp near me" and "private ultrasound london".

Wait: nobody searches for a private GP in London?

Here's the twist…

"gp private near me", a word-order variant, shows up in the London Doctors Clinic data with 14,800 searches a month.

Anyone doing keyword research for a medical practice from Keyword Planner alone reaches the wrong conclusion.

A SERP with a local pack, eight clinics and a directory does not exist without demand.

Keyword	Searches/month	CPC
dermatologist near me	33,100	£2.33
dermatologist london	9,900	£4.86
private doctor near me	5,400	£4.19
book gp appointment online	5,400	£2.29
medical centre near me	4,400	£1.90
private clinic near me	2,400	£4.48
trichologist near me	2,400	£3.60
nutritionist london	1,000	£3.74
private clinic london	880	£9.90
cardiologist london	720	£7.73
orthopaedic surgeon london	210	£4.43
sports medicine doctor london	170	£4.57
private cardiologist london, private dermatologist london, private gp london, private gp near me, private blood test london, private ultrasound london	no data	no data

Three sources that work instead of Keyword Planner

1. The queries report in Google Search Console, with the real searches the site already appears for.
2. The reception desk: the words patients use to ask for a treatment on the phone are the queries.
3. The related searches on the SERP. For "cardiologist london" they include "Cardiologist London NHS", "Private cardiologist london prices", "Best private cardiologist London" and "Cardiologist London Harley Street".

Spot the pattern?

Price, NHS versus private, and location. Nothing else.

The same gap appears outside London: for many "private + specialty + town" combinations the planner has no figure at all, and demand has to be estimated from Search Console and from the phone.

The pages that rank: treatment, location, clinician (most sites have only one)

A medical website that ranks has three types of page: one for each bookable service, one for each physical site and one for each doctor, all linked together. Almost every site we audit has only one, usually the list of specialties.

The treatment page

One page for each bookable service: cardiology consultation, echocardiogram, dermatology consultation, abdominal ultrasound.

Not a "Cardiology" page with a list.

So what goes on it?

The People also ask boxes tell you:

• for "cardiologist london": "How much does it cost to see a cardiologist in the UK?" and "Can you go straight to a cardiologist?";
• for "private gp london", three out of four are about price: "How much does a private GP cost in London?", "How much is private GP per month?", "Is it worth paying for a private GP?".

Some clinics already answer in the snippet: Mayfield Clinic shows "30 minutes | £220" in the result for "private gp london".

The page that ranks says:

• what's included and how long it takes;
• how much it costs;
• how to prepare and when the report arrives;
• at which site and who performs it;
• whether a GP referral is needed, and how to book.

That's the kind of information AI Overviews pull from treatment pages.

Two lines and a phone number are not enough.

The location page

One page for each physical site: address, opening hours, parking or nearest station, services available there and clinicians who work there.

It captures "brand + place" and "clinic + neighbourhood" searches.

London Doctors Clinic's location pages rank first for "london doctors clinic canary wharf", "london bridge" and "kings cross", and second for "clinic victoria" (2,900 searches a month).

We've applied the same principle, one page per territory, outside healthcare. For Sunscape, which looks for land for solar plants all over Italy, we built an SEO architecture by region: from March 2026, in six months, first-page keywords went from 1 to 38 out of 143 and organic impressions from 329 to 5,809 a month.

If you have one site, this page is the home page.

The clinician page

One page for each doctor: name, specialty, GMC number, services, sites and days.

It does two jobs:

• it captures searches for the name (patients look up the consultant their GP referred them to);
• it gives the treatment page a verifiable author, which is what Google asks of health content in its quality rater guidelines.

We explain it in what E-E-A-T means and how to demonstrate it.

How they link together

• the treatment page links to the sites where it's offered and the doctors who perform it;
• the location page links to its services;
• the doctor links to theirs.

This network of internal links is the structure of the site. Schema markup (MedicalClinic, Physician, MedicalProcedure) makes it readable for Google and AI assistants.

Recommended reading: how to build an SEO-friendly website.

Google Business Profile: the listing counts as much as the site (and 105 reviews can be enough)

In a healthcare local pack, proximity to the searcher, the listing's category and how well the name and services match the query come first, then reviews. On "cardiologist london" Heart Health at Welbeck is in with 105 reviews and 4.8, alongside two individual cardiologists with 379 and 238.

Only 105?

Yes. A clinic with 40 genuine reviews and the right category beats one with 300 and the generic category "Doctor".

The working rules, drawn from the listings we see in the pack:

• one listing for each site, with the correct primary category ("Medical clinic", "Private hospital", "Medical diagnostic imaging centre") and secondary categories for the specialties;
• a working booking link: it turns the listing into an appointment without going through the site;
• services added as "services", with the price where you choose to show it;
• replies to every review, never mentioning the treatment received or any clinical detail;
• photos of the rooms and entrances, not stock images.

And does it matter how many listings you have?

It does. On "private gp london" Private GP Surgery takes two of the three local pack slots with two separate sites, each with its own listing.

Recommended reading: how to optimise your Google Business Profile and the local SEO guide.

What you can write: GMC guidance and the CAP Code (the price yes, "the best" no)

On a UK clinic's website you can publish services, prices, locations and clinicians, but every claim must be accurate, provable and must not exploit the patient's vulnerability. That comes from the GMC's Good medical practice and from the CAP Code, and medical SEO has to start there.

Restaurant marketing doesn't have this constraint.

GMC, Good medical practice

The GMC's Good medical practice, domain 4:

• paragraph 89: any information a doctor communicates must be accurate, not false or misleading; doctors must take reasonable steps to check it, must not leave out relevant information, must not minimise risks of harm and must not present opinion as established fact;
• paragraph 90, which covers advertising your services: you must declare conflicts of interest and must not exploit people's vulnerability or lack of medical knowledge.

CAP Code, section 12

Section 12 of the CAP Code is what the Advertising Standards Authority applies to marketing for treatments and health services, websites included:

• rule 12.1: objective claims must be backed by evidence, where relevant trials on people;
• rule 12.2: marketers must not discourage essential treatment for conditions that need medical supervision;
• claims that a treatment is guaranteed to work or free of side effects need proof, and in practice are very hard to support.

Complaints go to the ASA, which publishes its rulings.

Translated into web pages

You can write: "cardiology consultation with ECG, 45 minutes, £250, report the same day, Dr Smith, our Harley Street clinic". That's information.

You shouldn't write: "London's best heart clinic", "unbeatable prices", "guaranteed results", "offer ends this month".

And the price?

You can show it, and for SEO it pays to. But it should be the full cost, not a base fee with extras added later.

Glowing testimonials on treatment pages are what we see challenged most often. Google reviews written by patients are a different matter.

We're not lawyers: on borderline cases the word belongs to the regulator and to your professional body. Clinics in England also register with the CQC; Scotland, Wales and Northern Ireland have their own regulators (Healthcare Improvement Scotland, Healthcare Inspectorate Wales, RQIA).

For SEO the constraint helps. The pages Google and AI Overviews reward are informative ones: what's included, how much it costs, who does it. They coincide with what the rules allow.

AI Overviews: how to get a treatment page cited (three things, one week)

To get a treatment page cited in an AI Overview you need a direct answer sentence under each H2, a medical author with name and GMC number, and a genuine last-updated date. AI Overviews on treatment searches cite treatment pages, not home pages.

The AI Overview on "cardiologist london" loads asynchronously and the crawler didn't return its sources.

So how do we know what gets cited?

From the equivalent Italian SERP we analysed ("visita cardiologica milano"): the AI Overview cited no home pages. It cited treatment pages from three hospital groups, plus one clinic's Facebook profile.

Pages that answer "how much, what's included, where" with full sentences and numbers. Not with a list of specialties.

The three things, doable in a week for your ten most requested services:

1. a direct answer sentence under each H2: "A cardiology consultation at clinic X costs £250 and includes a resting ECG and a consultation with the cardiologist";
2. a medical author with name and GMC number;
3. a genuine last-updated date.

We list the rest in what determines presence in AI Overviews.

Is there a flip side?

Yes. On condition searches ("thyroiditis symptoms", "what to do about high cholesterol") the AI Overview answers by itself and clicks drop.

These are the searches that bring Bupa millions of visits. And one more reason a clinic shouldn't target them.

Treatment searches with a town still get clicked, because they end in a booking.

Recommended reading: how AI Overviews work and our SEO for AI search service.

How to measure: bookings, not visits (Bupa has 3.3 million visits, and that's not the point)

The KPI for a medical practice is the number of bookings and calls from organic search, broken down by page. Traffic is a misleading number: Bupa has an estimated 3.3 million visits a month and most of them aren't looking for an appointment.

In practice, three events:

• the booking form;
• taps on the phone number on mobile;
• clicks to the external booking system.

Only three? Why not count scrolls and time on page too?

No. And we say that from direct experience.

In the first full month of the Google Ads account of Valore Solare, a solar installer in Italy (October 2025), the recorded "conversions" were 15,267 against 5,467 clicks. They were technical events: the real enquiries were 9.

Measure the wrong thing and you optimise the wrong thing. In SEO as in Ads.

In Google Search Console the queries report filtered by page shows which services bring clicks and which are stuck in positions 8-20, where the work pays off first. The AI features report, available since June 2026, shows which pages appear in AI Overviews.

One more constraint: health data

An event called "booking_oncology_consultation" tied to a user identifier is health data.

Better a generic "booking" event with the page as a dimension.

It's the same principle Google applies to advertising, and we cover it in the companion article.

Recommended reading: which SEO KPIs to track.

SEO or Google Ads for a medical practice? (both, in sequence)

For a medical practice the answer is both, in sequence: Google Ads brings bookings from day one while the treatment pages climb, and SEO takes months but then has no cost per click. The auction isn't cheap, though: Google Ads data puts competition at "high" on "cardiologist london" and "private doctor near me".

In the two medical SERPs the crawler found no ads, and in the DataForSEO database Bupa has 3 paid keywords and London Doctors Clinic 2.

So the big players aren't advertising?

Not necessarily. Absent ads in a single national snapshot often mean campaigns targeted by postcode or schedule that a UK-wide crawler doesn't see.

And the CPCs say the auction is real:

• £7.73 on "cardiologist london";
• £4.19 on "private doctor near me";
• £9.90 on "private clinic london".

One of our own cases, outside healthcare but with the same structure as a clinic: Contesini has two offices in the province of Mantua, in Italy, and had never used Google Ads. With campaigns focused on the area around the two offices: 299 enquiries at €7.20 each, 68% of them workable.

The sequence:

• Google Ads brings bookings from day one on "treatment + town" while the treatment pages climb, then stays on the high-value services;
• SEO takes months, but has no cost per click and doesn't switch off.

How to set up the paid side is in Google Ads for medical practices and clinics; how the two channels help each other, in SEO and Google Ads together.

For paid search in mental health there are specific constraints: Google's personalised advertising policy doesn't allow remarketing based on health, and professional guidance limits what ad copy can say. We explain them, with CPCs and budgets, in Google Ads for psychologists.

Frequently asked questions

Recommended reading: how to split marketing channels in a healthcare practice.

Where to start (ten services, one phone)

Not a blog on BMI and electrolytes, but a page for every service.

Not 379 reviews, but the right category: Heart Health at Welbeck is in the pack with 105.

Not "London's best", but "£250, 45 minutes, report the same day".

Not visits, but bookings and calls per page.

Now, two concrete things.

One: take the ten services your reception books most often and search for each on Google with the name of your town, on your phone. Note whether the AI Overview appears, whether there's a local pack, who's in it, and whether your site has a dedicated page with price, duration and clinician.

Two: count the noes. In the clinics we see, it's eight out of ten. Those eight pages are the first quarter of work, and writing them needs neither a blog nor an advertising budget.

If you'd like to look at the SERPs for your services together, book a strategy consultation: we do the first analysis on your own Search Console data.

See you next time!