01
Executive thesis
A medical website is decision-support infrastructure that happens to have a visual layer. It exists so that a person under some degree of stress can determine relevance, assess credibility, establish access and take a valid next step. Everything else is secondary to those four outcomes.
The industry approves these projects in the wrong order. Creative direction is signed off first, information architecture is fitted around it and content is written last, usually by whoever has capacity. By the time anyone asks where insurance acceptance goes, the design has no place for it. The site launches, stakeholders are pleased, and enquiry volume does not move.
The decision this article supports is what to approve and when. Six jobs must be solved before creative direction is signed off, because each one imposes structural requirements that cannot be retrofitted into a finished design.
02
Numbers that frame the issue
| Figure | What it tells us | Type | Source, date and geography |
|---|---|---|---|
| WCAG 2.2 Level AA sets a minimum target size of 24 by 24 CSS pixels, with defined exceptions | Tap targets are a measurable conformance criterion, not a design preference | Official accessibility standard | W3C, WCAG 2.2, current standard, global |
| WCAG requires labels or instructions identifying expected input for form controls | Booking and enquiry forms carry accessibility obligations independent of aesthetics | Official accessibility guidance | W3C, labels or instructions, current guidance, global |
Both criteria are single components of a broader standard. Meeting them does not constitute conformance, and the full success criteria including exceptions have to be applied rather than the numbers alone.
03
Why attractive healthcare websites still fail patients
Three patterns account for most of it.
The site is organised around the organization. Navigation mirrors the internal structure: departments, centres of excellence, about us, our vision. Patients do not think in departments. They think in problems. A person with chest pain does not know whether they need cardiology, internal medicine or the emergency department, and a menu built from the org chart asks them to make a clinical judgement before they can proceed.
Persuasion is deployed before relevance is established. The homepage opens with a statement about commitment to excellence and a photograph of a smiling clinician. Neither tells the visitor whether this provider treats their problem. Persuasion applied before relevance is noise, and on a small screen it is noise occupying the space where the answer should be.
Access information is treated as administrative. Insurance acceptance, referral requirements, opening hours, which services run at which site, what a first appointment costs. These are the questions that determine whether an interested person becomes a patient, and they are routinely buried in a footer, a PDF or a call centre. Access information is part of the value proposition, not a compliance detail.
None of the three is a visual problem. All three survive a redesign intact.
04
The six jobs a medical website must complete
The DEMA Six Jobs of a Medical Website define what must be solvable before creative direction is approved. Each job has a structural consequence, which is the reason for the sequencing.
Job 1. Establish relevance. The visitor must determine within seconds whether this provider handles their problem. Structural consequence: the site needs entry points organised by condition and symptom, not only by department.
Job 2. Demonstrate specific credibility. The visitor must find evidence that this organization and this named practitioner handle their specific case. Structural consequence: practitioner profiles need depth and must be linked from services, not buried in a directory.
Job 3. Establish access feasibility. The visitor must determine whether attending is realistic: location, hours, insurance, referral, waiting time. Structural consequence: access data needs a home on service and location pages, which means design must accommodate dense factual content.
Job 4. Enable a valid next step. Every page must offer an action appropriate to where the visitor is in their decision. Structural consequence: several action types are needed, not one universal book-now button, since a person still identifying their problem is not ready to book.
Job 5. Support the appointment already made. Preparation instructions, what to bring, where to park, how to reschedule. Structural consequence: content must exist for people who are not acquisition targets, which is why it is usually missing.
Job 6. Remain usable under stress and constraint. On a phone, on a poor connection, one-handed, in either language, by someone anxious or in pain, or by someone using assistive technology. Structural consequence: accessibility and performance are entry requirements rather than a late-stage audit.
A design that cannot accommodate all six is not a design problem to be solved later. It is the wrong design.
05
Decision support versus promotional design
The distinction is concrete rather than philosophical.
| Promotional approach | Decision-support approach |
|---|---|
| Homepage hero states organizational values | Homepage hero routes by problem, specialty and location |
| Service page describes the department | Service page states conditions treated, who delivers it, where and how to access it |
| Practitioner profile lists qualifications | Practitioner profile states scope, conditions treated and languages, with direct booking |
| One book-now action repeated site-wide | Action matched to decision stage: learn, compare, check access, book |
| Insurance information in a PDF | Insurance acceptance stated on the pages where the decision happens |
| Photography carries the emotional message | Photography supports the message; the text carries it |
Promotional design is not wrong in itself. It is wrong as a first layer. Once a visitor has established relevance, credibility and access, a strong visual identity does meaningful work in differentiating the provider. Applied before those things, it fills the space they needed.
06
The evidence every core page must surface
| Page type | Must answer | Usually missing |
|---|---|---|
| Homepage | What do you treat, where and how do I start? | Any route organised by patient problem |
| Service page | Who is this for, what does it involve, who delivers it, where, how do I access it? | Named practitioners, sites offering it, insurance position |
| Condition page | What is this, how serious, which specialty, what happens next here? | A link to the service that treats it |
| Practitioner profile | Does this doctor treat my case, in my language, and can I book with them? | Stated scope and direct booking |
| Location page | What runs here, who works here, when is it open, how do I get in? | Service list specific to that site |
| Booking or contact | What happens after I submit, and when will I hear back? | Confirmation of what to expect |
The most consequential gap in the table is the condition-page link to a service. It is the point where clinical information becomes commercially useful, and it is missing from most healthcare sites. The architecture behind it is set out in The Healthcare Search Architecture: Symptoms, Conditions, Procedures, Doctors and Locations.
07
A homepage and service-page teardown
The following teardown is illustrative. It describes a composite pattern rather than a specific organization, and the scoring is demonstrative rather than measured.
Composite homepage. Full-width video of a corridor. Headline referencing commitment to patient-centred care. Three cards: About Us, Our Doctors, Careers. Below the fold, a news carousel. Primary navigation by department, eleven items. A book-appointment button in the top right leading to a login screen.
Assessed against the six jobs: relevance fails, because nothing names a condition. Credibility is generic rather than specific. Access is absent. The next step is available but only valid for someone who has already decided and already holds an account. Job five is absent. Job six is untested but the video suggests a performance problem on mobile.
Score: 1 of 6.
Composite service page. Title, two paragraphs describing the department, a stock image, a phone number.
Relevance is partial, since it names the specialty but no conditions. Credibility is absent, with no practitioners named. Access is a phone number only, with no hours, no insurance position and no statement of which sites offer the service. The next step is one channel, unsupported.
Score: 1 of 6.
What the redesign brief should have said. Before any visual direction: the homepage must let a visitor reach the right service by problem within two interactions, and every service page must name conditions treated, list practitioners, state sites and state insurance position. These are content and structure requirements. They can be written before a designer is briefed, and they constrain the design in ways that are cheap now and expensive later.
08
The Medical Website Decision Score
Score each job from 0 to 2. Zero means the job is not addressed, one means partially, two means a visitor can complete it unaided on a phone in either language.
| Job | Diagnostic question | Score |
|---|---|---|
| Relevance | Can a visitor reach the right service by describing their problem rather than naming a department? | |
| Specific credibility | Does every service name the practitioners who deliver it, with stated scope? | |
| Access feasibility | Are hours, sites, insurance position and referral requirements stated where the decision happens? | |
| Valid next step | Does each page offer an action appropriate to that decision stage? | |
| Post-booking support | Does preparation, parking and rescheduling content exist? | |
| Usability under constraint | Does the site meet accessibility criteria and perform on a mid-range phone, in Arabic and English? |
Interpretation: 0 to 4 means the site is a brochure. 5 to 8 means it supports decisions partially, usually failing on access. 9 to 12 means the structure is sound and effort belongs in depth and content quality.
Run this before commissioning a redesign. A site scoring 7 rarely needs rebuilding. It needs the missing content, which costs a fraction of a rebuild. Diagnosing that distinction is the subject of Why Healthcare Websites Get Traffic but No Inquiries.
09
A redesign prioritization plan
Fix now
- Publish insurance acceptance and referral requirements as pages on the site
- Add practitioner names, with links, to every service page
- State which services run at which sites, on both service and location pages
- Replace the single universal booking action with actions matched to decision stage
- Check tap target sizes and form labels against WCAG criteria
Build next
- Add condition-level entry points and link each to the service that treats it
- Build practitioner profiles with stated scope, languages and direct booking
- Create preparation and post-booking content for the highest-volume services
- Define approval criteria for the next design phase that include content and access requirements, not only visual direction
Measure continuously
- Decision Score by page type, reviewed at each content release
- Exit rate on access-related pages
- Service-page to booking-start rate
- Accessibility conformance as a standing check rather than a launch gate
Mobile performance and constraint usability get fuller treatment in Mobile Access Is Healthcare Access: A Practical UX Standard, and the practitioner layer in The Doctor Profile as a Conversion Page: 12 Signals Patients Need.
10
Audit methodology and limitations
What this article is. A position paper proposing six nonvisual jobs as approval criteria for healthcare website projects, with a scoring instrument. The six jobs are an editorial contribution drawn from recurring failure patterns.
What it is not. The teardown is a composite illustration, not an audit of any organization. The scores are demonstrative. No DEMA website audit benchmark has been published and no completeness rate should be inferred.
Standard limitations. The two WCAG criteria cited are individual success criteria. Meeting them is not conformance. Full accessibility assessment requires the complete standard, appropriate testing and, in most cases, specialist review.
Where review is required. Condition and service content requires clinical review before publication. Booking flows that collect patient information require privacy review under the Saudi Personal Data Protection Law. This article is not legal advice.
Planned research. DEMA has scoped a medical website decision-support audit across Saudi Arabia and the GCC. It is not complete, and this article contains no empirical audit findings.
Sources.
11
About to approve a redesign?
The most expensive healthcare website mistake is approving creative direction before anyone has established what the site must let a patient do. A Healthcare Growth Diagnosis scores your current site against the six jobs, in Arabic and English, and usually shows that what looks like a rebuild is a content problem.
Request a Healthcare Growth Diagnosis