Send the same one-paragraph brief to three agencies and you will get back three quotes that do not resemble each other. Not slightly different — different by a multiple.
Everyone assumes one of them is lying. Usually none of them is. They have each read the brief and imagined a different piece of work, because "a website for our clinic" describes an outcome rather than a scope, and the scope is the entire price.
The useful question is not what a medical website costs. It is what you are being quoted for.
The six things inside every quote
Every medical website quote, whatever the number, is really six separate jobs bundled into one line. When two quotes differ wildly, it is because one of them includes jobs the other assumed you would do yourself.
Information architecture. What pages exist and how they relate. For a single-doctor practice this is an afternoon. For a hospital with departments, doctors, procedures, locations and international patients, it is the largest thinking task in the project and the one that determines whether the site can be found in search at all.
Content. Who writes the page about your knee replacement pathway. If the answer is "the client will supply", the quote is lower and the project will stall, because clinics almost never supply. Medical content also needs a clinician to review it, and that review has to be scheduled around clinic hours.
Design. Whether you are buying a theme configured to your colours or an identity built for your practice. Both are legitimate. They are not the same purchase.
Build and integration. The site itself, plus everything it has to talk to: appointment booking, the practice management or hospital system, WhatsApp, a CRM, payments, a patient portal. Integrations are where quotes diverge most, because each one is a separate piece of engineering and most briefs mention none of them.
Compliance and accessibility. Consent handling, the privacy notice, cookie behaviour, and accessibility conformance. In the US this carries direct legal exposure — HIPAA where patient data is involved, and ADA-related litigation over inaccessible sites is an established pattern. In India the DPDP Act 2023 and the DPDP Rules 2025 govern consent and notice. In the UAE and Saudi Arabia the local data laws apply instead. Whichever market you are in, WCAG 2.2 AA is the standard worth building to, and retrofitting it later costs more than building to it.
Migration. The part nobody quotes and everybody needs: redirect mapping from the old URLs to the new ones. Skip it and you lose the rankings the old site had. This single omission is responsible for more post-launch panic than any design decision.
Four scopes, described by what they contain
Rather than pretending a price exists in the abstract, it is more useful to recognise which of these four you are actually buying.
A presence site. A handful of pages for a solo practitioner: who you are, what you treat, where you are, how to book. No integrations beyond a form. Correct choice for a new practice that needs to exist online before it needs to compete.
A conversion-built clinic site. The same thing, engineered for the patient decision: a page per condition and per procedure, a booking path that works on a phone in three taps, doctor profiles with real credentials, location pages, schema, speed. This is the tier most clinics actually need and the tier most of them under-buy.
A multi-location site. Everything above plus a structure that can hold five, twenty or eighty locations without duplicating content, with per-location pages that are genuinely different from each other and a content model your team can add to without a developer.
A hospital platform. Departments, service lines, hundreds of doctor profiles, international patient journeys, multiple languages, an events and news operation, and integration with systems that were not designed to be integrated with. This is software, not a website, and it is priced like software.
The mistake that costs the most is buying tier one, discovering within a year that you needed tier two, and paying for both. If you intend to compete in search, start at tier two.
The line items that turn up later
A quote that looks cheap usually excludes a predictable set of things. Ask about each one explicitly before you compare:
- Copywriting, and how many rounds of clinician review are included
- Photography of the actual premises and the actual doctors — stock photography of a smiling model in a white coat is visible to patients and costs you trust
- Translation, and whether translated pages get their own keyword research or just a rendering of the English
- Booking or EMR integration, named system by named system
- The redirect map and the post-launch crawl to catch what broke
- Analytics, consent management and conversion tracking set up and verified, not just installed
- Accessibility testing against WCAG 2.2 AA
- Hosting, SSL, backups, security updates and who applies them
- Who can edit the site afterwards, and whether editing needs the agency
That last one deserves its own sentence. A site your marketing coordinator cannot update is a site that will be out of date within six months, and every update will be billed.
Template versus custom is the wrong argument
The real question is not whether the site is built on a template. It is what happens when your practice changes.
You add a specialty. You open a second location. A doctor leaves and another joins. You want a page for a procedure you did not offer last year. If those changes are a form your team fills in, the platform was built properly. If each one is a quote from the agency, you did not buy a website — you bought a dependency.
Ask for a demonstration of the admin, not a description of it. Ask to add a doctor while you watch.
