SEO for Doctors: How Patients Actually Find a Practice

A practice that has held the same corner for fifteen years has a full book and no obvious reason to be found. That holds until it does not: a principal cuts back, a partner leaves, a new practice opens two blocks away with parking and online booking.

The demand side is not the problem. The RACGP's Health of the Nation 2025 report, the college's annual report on general practice, puts the market at 172 million health services delivered to 22.6 million patients in a year. The patients exist. What decides whether they arrive at your door is whether your site answers the question they typed into a phone.

Thirteen years of this work has shown me the same three faults on practice sites again and again. No page for the thing patients actually search for. An appointments path that dead-ends in a phone number nobody answers. A new-patient position that is either stale or absent. What follows is how I would fix those three, in the order I would fix them.

One boundary, stated once. This is marketing guidance. It is not medical advice, not legal advice, not a view on registration or advertising obligations, and nothing here promises a ranking, a patient volume or a clinical outcome.

The Demand Is Already There, and It Is Not Going to Your Site

What the Medicare data shows about how often Australians see a doctor

Start with the size of the pool, because it reframes the problem. The AIHW's trends page for Medicare-subsidised GP, allied health and specialist care, which covers 2017-18 to 2024-25, states: "More than 8 in 10 Australians had at least one Medicare-subsidised GP attendance."

The same page shows how the visit rate has moved since: "GP attendances increased each year between 2017-18 and 2021-22 (about 627 per 100 people to 736 per 100), before decreasing to 615 per 100 people in 2024-25." That is the sentence worth sitting with. Visits per person climbed for four years and then fell back. Fewer visits per person means every patient decision carries more weight for the practice, and the competition for the same visit is sharper than it was in 2021.

A practice is also not competing for a single transaction, and the long arc makes that plain. The AIHW summary puts it this way: "From Medicare's inception in 1984 until 2025, the yearly GP services attendance rate has increased from 3.8 to 6.1 visits per person (AIHW 2026)." Attendance is a repeat behaviour, which is why a findable page is worth more here than in most service businesses. Section 8 puts that into arithmetic.

Why an established patient base hides a weak website

A full book is not evidence that the site works, because the book was built by referral, location and time. None of those tell you what happens when a family moves into the area and searches, or when a patient needs a service you offer and have never published.

The practices I worry about are not the ones with an empty diary. They are the ones with a solid diary and no way to tell where the next ten years of patients will come from. Fees, new-patient status and service mix are the three things a new arrival needs, and the three things a practice rarely writes down.

The patient who is not yours yet, and what they type

That patient is not searching for a practice. They are searching for a problem, a suburb or a constraint: a GP taking new patients, a skin check near me, a practice open on Saturday, a doctor who speaks Mandarin, somewhere that bulk bills. Each of those is a different question, and on most practice sites none of them has a page. The site is organised around the practice's own structure, which is our team, our history, our services. The patient is organised around their afternoon.

Three Faults That Appear on Almost Every Practice Site

The homepage that describes the practice instead of the problem

The typical homepage opens with the year the practice was founded and a photograph of the building. A patient has two questions at that point: can I get in, and what will it cost me. Neither is answered above the fold, and the second is usually not answered anywhere on the site.

The fix is not a redesign. It is two lines near the top: whether the practice is taking new patients, and what a standard consultation costs. Everything else on the page can stay where it is.

The appointment path that dead-ends in a phone number

A booking button that opens a phone dialler during business hours is not a booking path, it is a queue. The patient who searched at 9pm, on a phone, with a child asleep, calls nobody. Count the steps between "I want an appointment" and a confirmed time. Three is workable. Six is a funnel with a hole in it.

The stale "accepting new patients" position

Almost every practice site has a line about new patients somewhere, and it is usually out of date, buried on a contact page, or both. For a patient who has just moved into the area, that sentence is the most consequential one on the site, and it is the sentence nobody owns. Give it to one person, put a review date on it, and correct it the week it changes.

What a Patient Is Actually Deciding

Before the pages, the decision. Four questions run through a patient's head in roughly this order, and each one belongs to a page rather than to a brochure.

Can I get an appointment, and how soon

The first question is availability, and the honest version of it. "Same day" means same day, and "usually within a week" is a real answer that filters out the people who needed today. A practice that only ever says "contact us" makes the patient do the work, and the patient is holding a phone with a search box in it.

Where is it, and can I get there or get online

Parking, transport, accessibility, and whether a consultation can happen by video. A patient who cannot park will not book, however good the doctor is, and that detail belongs in a line of text rather than only on a map.

What will it cost me, and is bulk billing available

This is where the sector data cuts in. The RACGP's report, drawing on ABS data, records that "the proportion of people delaying GP care due to cost has increased in the previous 12 months from 7% to 8.8%". That is a population figure rather than a claim about your patients, and it points one way: somebody deciding where to go is now more likely to be weighing cost before booking. A fees page carrying the standard consultation fee, what bulk billing covers and what it does not costs nothing to write, and it answers the question a patient is least willing to ask on the phone.

Is this person qualified to help with my thing

The last question is the one practices answer worst, because they answer it with a directory listing. A name, a qualification and a photograph is what a directory already provides. Credibility that earns an appointment is more specific, and section 5 deals with it.

One counterweight before the marketing argument gets too confident. Access in Australia is generally good, and the RACGP reports: "In 2023-24, 99% of people reported that they were able to see a GP when they needed general practice care, and more than half of the population could access urgent medical care from a GP within 24 hours." Most people can get care. That makes the practice that is hard to book the exception, and exceptions get noticed and mentioned. It does not make a website optional. It means the standard to clear is competence rather than scarcity.

The Three Pages a Practice Genuinely Needs

The condition or service page for what you are actually searched for

Pick the two services that fill your week and write a real page for each. Not a paragraph in a services list. A page that says what the service is, who it suits, what happens at the appointment, how long it takes, what it costs and how to book it. That is the page that answers a search question, and it is the page a patient forwards to a partner before booking.

Two pages done properly beat twelve done thinly, and the second half of that sentence is the most common mistake I see. A practice with a page for every service it technically offers ends up with twelve pages that say the same thing at the same length with the name changed.

The appointment and access page, including fees and bulk billing

One page that answers four things: how to book, where you are and where to park, what to bring, and what it costs, including what happens if a patient needs to cancel. Link it from the header, from every service page and from the booking confirmation. If a patient has to hunt for the fee, they book somewhere the fee is visible.

The practitioner page that carries credentials, interests and languages

One page per practitioner, not a card in a grid, and section 5 covers what belongs on it. The commercial reason it matters is simple. When a patient's regular doctor is away, or a family needs a second appointment type, the practitioner page is what keeps the appointment inside the practice instead of sending it down the road.

I have done this work in healthcare before, and the healthcare case study pages are where to look if you would rather see the shape of the work than read a description of it.

Writing Practitioner Pages That Do Real Work

What a patient wants to know and what a directory already tells them

A directory lists your name, your qualifications and your suburb. If your page repeats that and stops, the page has no reason to exist. The question behind the click is narrower: does this person deal with what I have, and will I be comfortable in the room?

Credentials, special interests, hospitals, languages spoken

What the practitioner actually treats, week to week. Special interests, stated plainly. Where they trained and where they have worked. Which hospitals they attend. Languages they consult in, which for a large share of patients settles the booking on its own. All of it is verifiable, and none of it is a clinical claim.

Why a generic team grid is the weakest version of this page

A grid of faces and job titles gives a patient no way to choose and no reason to come back to the site. It also wastes the strongest asset a practice owns, which is the practitioners themselves. Most industries would pay for that credibility and cannot manufacture it. A medical practice has it and usually publishes it as a photograph.

The Appointment Path Is the Conversion

Three routes in: online booking, phone, and a form

All three should exist, and all three should be visible from the same place. Online booking suits the patient at 9pm. The phone suits the patient with a complicated problem. A form suits the patient who does not want to talk to anyone yet, and if the practice does not want forms, it should say so rather than leave one on the site going nowhere.

What breaks when the booking link is third-party and slow

The usual failure is a booking button that leaves the site, loads slowly on mobile and asks the patient to create an account before showing a time. Each of those steps loses people, and none of them shows up in a monthly report. Test the path on a phone, on mobile data, as a patient would, and time it. If it takes more than a minute to reach a confirmed appointment, that is the first thing to fix, ahead of any new page.

Answering the phone as a marketing decision

The cheapest conversion improvement in most practices is not on the website. It is the phone being answered, or a message being returned inside an hour. A missed call from a new patient is a booking that went to whichever practice picked up. If you want to know what your marketing is worth, count the calls that rang out last week.

Being Findable Locally Without Buying Every Ad

The suburb and service combination that patients actually type

Patients combine a service with a place. The condition or the appointment type, next to the suburb name or the station they commute from. The practical consequence is that a page built around a service, with a clear statement of where it is delivered, matches the words in a patient's head far better than a page about the practice's philosophy. That is not a technique. It is a description of what the reader typed.

Why one strong service page beats twelve thin suburb pages

A dozen pages with the suburb name changed are the most common form of wasted work in this sector. They compete with each other, they carry no information, and a patient who lands on one can tell within a paragraph that it was not written for them. One service page, written properly and supported by clear location detail, does more for the same effort and does not put the practice's credibility at risk.

Where paid search fits, and what it should not be asked to fix

Paid search buys a place in the queue while the pages catch up, and it is worth running on the two or three services where a patient is worth more than the click. It should not be asked to fix a site that cannot answer the fee question, because the click lands on the same dead end the organic visitor found. The paid route is a budget decision rather than a technical one. If the honest problem is that nearby patients cannot find the practice at all, that is the local side of the work and a separate conversation from the pages.

The Worked Example: One Practice, Three Pages, One Quarter

Listing what you already publish that nobody reads

Before writing anything, list every page on the site and mark what each one is for. Most practices find two things: pages nobody visits, and questions patients ask on the phone that have no page at all. The second list is the valuable one, and it usually takes an afternoon to assemble from the front desk.

If you want a second opinion on that inventory rather than a rebuild, a structured review of what the practice is already publishing is where I would start.

Choosing the two services worth pages

Rank the services by how often patients ask for them and how much the practice wants that work. The top two become pages this quarter. Everything else waits. A quarter is long enough to write two good pages and short enough that you will actually finish them.

The four numbers to record before you start

The arithmetic only means anything with your inputs, so I am leaving the values to you. Record these four before the first page goes live.

• Services the practice provides in a week, counted from your own booking system.
• The share of first-time patients who say they found you through search, asked at the desk if the booking system cannot tell you.
• What one visit is worth to the practice.
• The hours and the cost to write, build and publish the two pages.

Then one external input, and only one. The AIHW reports 6.1 visits per person a year. Multiply the value of one visit by that figure for an order-of-magnitude view of what a patient is worth across twelve months. It is a population average covering all Australians rather than a prediction for your patients, so treat the result as a scale check rather than a forecast.

The comparison is then simple. Two pages, with the hours and the cost attached, against the number of additional patients those pages would have to produce at that value to pay for themselves. Work out that break-even number first. It is usually small, and if it is not, the two pages you picked are the wrong two.

The published figures for the alternative are specific rather than a request to get in touch. The SEO consultant page states: "SEO retainers in Australia typically start around $3,000 a month, diagnostic projects start around $5,000, and hourly consulting runs $100 to $300." That is how I work through a practice's pages before anything is rebuilt. If the engagement question is the open one, the engagement models and what each one costs are published. If the budget question is the open one, work out what the practice can put behind this first.

For the paid comparison, Semrush puts the cost per click on "seo for doctors" in Australia at $18.45. That is a third-party estimate of what advertisers pay for the term, not the practice's own account data, and it is the reason I would not set a budget on a click price alone.

What to Measure, and What to Ignore

Appointments booked from the site, not sessions

The number that matters is appointments, from any route, and where the patient came from. Sessions tell you the site is reachable. They do not tell you whether it is doing anything.

The share of first-time patients who found you online

Ask at the desk for one month and write it down. Four seconds per patient, worth more than any dashboard, because it is the only number that connects the site to the diary. If the booking system can capture the source, better. If it cannot, the desk is the measurement instrument.

Why a ranking for a suburb name is not the goal

A ranking is a means, not an outcome. A practice can rank for its own suburb and still lose the patient who searched for a service it never wrote a page for. The AIHW figures show visits per person falling back, which means the visit is contested. The practice that wins the contest is the one that answers the specific question, not the one that owns a place name.

What Not to Do

Do not build a page per suburb with the name changed

It wastes the effort, competes with itself and reads as filler to the patient. If a suburb genuinely needs its own page, it needs its own information: access, hours, practitioners, services actually delivered there.

Do not publish clinical claims to win a search result

A search result is not worth a claim you cannot support. Describe what the practice does, who it suits and what happens at an appointment. Anything beyond that belongs to the clinical side of the business, not the marketing side.

Do not hand the website to whoever answers the phone fastest

The site carries fees, availability and credentials, so the person responsible for it should be the person who owns those facts. In most practices that is the practice manager, with the principal signing off. When it defaults to an agency or a relative who builds sites, nobody can update the fee page, and the fee page is the one patients need.

Frequently asked questions

How do patients actually find a new doctor?

Mostly through search and maps, and usually with a problem rather than a category. Someone moves suburb and searches for a GP near them; someone with a persistent symptom searches for the condition and the suburb together; someone who has been waiting too long searches for a practice that is accepting new patients. Each of those is a different page, and most practice sites answer none of them, because the site is organised around the practice's internal structure instead of the patient's question. Fix the three questions first: can I get in, where are you and what will it cost, and can you help with my thing.

We already have a full patient book. Why would we need this?

Because a full book and a full list of the right patients are different things. The attendance data shows visits per person fell back to 615 per 100 people in 2024-25 from a peak of 736, so patients are choosing where those visits go. A practice with no findable service pages, no fee information and no clear new-patient position loses the patient who moves into the area, the patient whose usual doctor retires, and the family that needs a second appointment type you do offer but never say you offer. You do not need more patients to justify this. You need the right ones to be able to find you.

Should each doctor have their own page?

Yes, and it should be more than a photo and a title, because that is what a directory listing already gives a patient. The page that earns an appointment explains what the practitioner actually treats, their special interests, where they trained and worked, which hospitals they attend, and the languages they consult in. That page also does something a team grid cannot: it gives a patient a reason to choose one practitioner over another inside the same practice, which is how a practice keeps a patient when their regular doctor is away.

How long does this take to work?

Longer than paid search and shorter than most people fear, and I will not put a number of months on it, because a practice in a dense suburb competing with a hospital group is a different job from a solo GP in a regional town. What I will say is that the first change that shows up is usually not traffic, it is the quality of the phone calls: fewer people asking where you are and what you charge, more people asking whether they can get in on Thursday. That shift is the leading indicator, and it usually appears before any ranking movement.

Is this different from what a marketing agency would do for a practice?

The mechanics are the same as any local service business; the judgement is not. A medical practice has a register of services it can and cannot claim, a fees structure patients genuinely need to know before booking, and a credibility asset that most industries would pay for, which is the practitioners themselves. What I would avoid is the pattern I see most often: a suburb page per postcode with the name swapped, and a blog of general health articles that compete with healthdirect and the hospitals rather than with other practices. Neither of those wins an appointment.

What should we measure first?

Appointments booked, from any route, and how the patient found you. That is it for the first quarter. If your booking system will not tell you the source, ask at the desk for one month and write it down: the answer takes four seconds per patient and it is worth more than any dashboard you could buy. Sessions, rankings and keyword counts come later, and only once you can answer the appointments question.

Where to Start This Month

Three things, and none of them needs a rebuild. Publish the fees and bulk billing page. Fix the appointment path on a phone, on mobile data. Ask every first-time patient at the desk how they found you, and write the answer down.

That is a fortnight of work by somebody already inside the practice, and it produces the three things the rest of the plan needs: a page that answers the cost question, a booking path that works at 9pm, and a baseline for judging whether anything changed.

If you would rather have the two pages chosen before you commit to writing them, send me your three most-booked services and I will tell you which pages I would write first.