01
Executive thesis
Booking friction is usually blamed on form length. Form length is the smallest part of it. The larger costs come from channel switching, account requirements imposed before value is delivered, unstated eligibility, and the silence between submitting a request and learning whether it worked.
The consequential distinction is between avoidable and legitimate friction. A hospital that requires a national ID for a procedure with regulatory identity requirements is not being careless. A hospital that demands the same ID before telling the patient whether the service is even offered at that site is. Any measurement that treats all friction as waste will produce recommendations the clinical and compliance teams correctly refuse.
The decision this article supports is where to spend limited engineering effort. Score the sequence, separate the avoidable friction from the necessary, and fix in order of cost.
02
Numbers that frame the issue
| Figure | What it tells us | Type | Source, date and geography |
|---|---|---|---|
| More than 16 million virtual appointments and medical consultations delivered in one year | Saudi patients complete digital health transactions at national scale, so digital booking capability is not the constraint | External evidence | Saudi Ministry of Health, 2025, Saudi Arabia |
| WCAG 2.2 sets a minimum target size of 24 by 24 CSS pixels at Level AA, with defined exceptions | Tap target size in booking flows is a measurable criterion | Official accessibility standard | W3C, WCAG 2.2, current standard, global |
| WCAG requires labels or instructions identifying expected input | Unlabelled or ambiguous booking fields are an accessibility failure, not only a usability one | Official accessibility guidance | W3C, labels or instructions, current guidance, global |
The Ministry of Health figure describes national virtual-care volume and does not establish private-sector booking completion rates. The accessibility criteria are individual components of a wider standard.
03
Why booking is an access system, not a form
The form is the last four percent of the sequence. Everything before it determines whether the patient reaches the form believing it will work.
Consider what a patient must resolve before the first field: is this service offered here, does this provider take my insurance, do I need a referral, is there availability this week, is this the right department, can I do this in Arabic. If the site answers none of these, the patient enters the form carrying six unresolved questions, and every field becomes a point at which they reconsider.
This is why form redesign projects so often produce disappointing results. The form was never the obstacle. It was the place where accumulated uncertainty finally became abandonment. The upstream drop points are covered in Search-to-Booking Leakage: Six Places Healthcare Providers Lose Qualified Patients.
Booking is also not finished at submission. A request that produces no confirmation, or a confirmation with no timeframe, leaves the patient uncertain whether they hold an appointment. Some call to check, which converts a digital transaction into a phone call and doubles the cost. Others book elsewhere.
04
The components of the Booking Friction Index
The index scores six components. Each is scored 0 to 5, where 0 is frictionless and 5 is severe, giving a total from 0 to 30. Higher is worse.
Component 1. Path length. The number of interactions from the service page to a submitted request. Count clicks, taps and page loads.
Component 2. Input burden. The number of fields required, weighted by difficulty. A name is easy. A policy number the patient must find in a drawer is not.
Component 3. Channel switching. Every forced move between channels. Website to phone, website to WhatsApp, web to app, app to a separate portal login. Each switch is a discrete opportunity to stop.
Component 4. Account requirement. Whether an account or login is demanded, at what point, and whether the reason is explained. A login required after the patient knows the service is available and appropriate is far cheaper than one required before.
Component 5. Eligibility uncertainty. Whether the patient can determine, before starting, that the service is offered at that site, that their insurance is accepted and whether a referral is needed.
Component 6. Confirmation latency and clarity. How long until confirmation, and whether it states what was booked, with whom, where and what happens next.
The scoring guide:
| Score | Meaning |
|---|---|
| 0 | No friction of this type |
| 1 to 2 | Minor, unlikely to cause abandonment alone |
| 3 | Noticeable, contributes to abandonment in combination |
| 4 | Significant, causes abandonment independently |
| 5 | Severe, blocks a meaningful share of patients outright |
05
Clicks, fields, channel switches and uncertainty
Two rules make the index useful rather than decorative.
Rule one: score the sequence, not the page. Friction accumulates. A five-field form is trivial in isolation and expensive as the fourth step after two channel switches and a login. Always measure from the service page, not from the booking page.
Rule two: classify every point of friction as avoidable or legitimate. Legitimate friction serves a clinical, regulatory or safety purpose that cannot be met another way. Avoidable friction serves an internal convenience, a legacy system or a data appetite that exceeds the need.
| Friction point | Usually legitimate | Usually avoidable |
|---|---|---|
| Identity verification for regulated services | Yes, where required | Not when demanded before service eligibility is known |
| Insurance details | Yes, before attendance | Yes, when demanded before the patient knows the insurer is accepted |
| Account creation | Sometimes, for records access | Usually, when required to make a first request |
| Referral confirmation | Yes, where clinically required | Not when the requirement is unstated until the final step |
| Full medical history at booking | Rarely | Usually, this belongs after the appointment exists |
| Phone-only booking for complex services | Sometimes | Not as the sole route for routine appointments |
The pattern in the right column is consistent. Most avoidable friction comes from asking for information before the patient has any reason to believe the effort will be rewarded. Sequencing is often the entire fix, with no reduction in what is ultimately collected.
06
Accessibility and bilingual friction
Two dimensions are usually excluded from friction discussions and both affect real patients.
Accessibility. Booking flows fail WCAG criteria in predictable ways: tap targets below the minimum size, date pickers that cannot be operated by keyboard, fields with placeholder text instead of labels, and error messages conveyed by colour alone. These are conformance issues with a measurable standard, and they exclude patients rather than merely inconveniencing them.
Bilingual parity. The Arabic and English booking paths must reach equivalent outcomes. Common failures: the Arabic path has fewer available services, the Arabic form reverts to English at the payment or confirmation step, the confirmation message arrives in English only, or the Arabic flow requires the patient to enter their name in Latin characters. Each of these is a friction point that only Arabic-speaking patients experience, which is why it survives testing conducted in English. Parity measurement is developed in [Arabic Healthcare Content Is Not Translation: The Bilingual Trust Parity Framework].
Score both dimensions as part of the index rather than as a separate audit. Friction experienced by a subset of patients is still friction.
07
A worked friction-score comparison
The following comparison is illustrative. The two flows are composites constructed to demonstrate the scoring method. The scores are not measured, not sampled and not a DEMA benchmark.
Flow A. Patient reaches a service page, taps Book, is required to create an account, verifies by SMS, completes eleven fields including insurance policy number, submits, receives no immediate confirmation and is told the clinic will call within two working days. Insurance acceptance is not stated anywhere.
Flow B. Patient reaches a service page that states which sites offer the service, which insurers are accepted and whether a referral is required. They tap Request appointment, complete four fields, submit, receive an immediate on-screen confirmation with a reference number and a message stating a call will follow within four hours. Account creation is offered afterwards, with the benefit explained.
| Component | Flow A | Flow B |
|---|---|---|
| Path length | 4 | 1 |
| Input burden | 4 | 1 |
| Channel switching | 3 | 1 |
| Account requirement | 5 | 0 |
| Eligibility uncertainty | 5 | 0 |
| Confirmation latency and clarity | 4 | 1 |
| Total | 25 | 4 |
Reading it. Flow B collects less information at the point of request, but it does not collect less information overall. It collects it after the patient has a reason to continue. The clinical and administrative requirements are identical. The sequence is not.
Decision implication. In Flow A the two most expensive components are account requirement and eligibility uncertainty, both scoring 5, and neither is a form problem. Publishing insurance acceptance on the service page and deferring account creation addresses 10 of the 25 points without touching the form. That is the cheapest available improvement and it requires no engineering.
Limitations. The scores are assigned by judgement against the scoring guide, which means two assessors may differ by a point on any component. The index is designed to compare flows and track change over time, not to produce a precise absolute value. Consistency of assessor and method matters more than the number itself.
08
How to reduce friction without compromising governance
Fix now
- State insurance acceptance, referral requirements and which sites offer each service on the service page
- Move account creation to after the request, and explain what it gives the patient
- Send an immediate confirmation containing what was requested, a reference and a stated response time
- Remove any field not needed to schedule the appointment
- Check tap targets, field labels and error messaging against WCAG criteria
Build next
- Score every booking path with the index, including the Arabic path scored separately
- Establish which data must be collected before attendance and which can follow, with clinical and compliance sign-off
- Bring the Arabic flow to parity end to end, including confirmation messages
- Instrument the flow so abandonment is visible by step rather than in aggregate
Measure continuously
- Booking Friction Index by path, rescored quarterly
- Abandonment rate by step, split by language and device
- Proportion of digital requests that become phone calls, which indicates confirmation failure
- Attendance rate against booked appointments, connecting friction to The Healthcare Acquisition Equation: From Click to Attended Appointment
Reducing friction is not the same as removing governance. Every change that touches identity, insurance or clinical eligibility should be agreed with the teams accountable for it. The index makes that conversation specific by attaching a cost to each requirement rather than arguing about the principle.
09
Methodology and limitations
What this article is. A scoring instrument for booking friction, built from observed patterns in healthcare booking flows and published accessibility criteria.
What it is not. The two flows compared are constructed composites. The scores demonstrate the method. No DEMA booking friction benchmark exists yet, so no industry average, no percentile and no target score is offered here.
Method limitations. Scoring involves judgement. The index compares flows and tracks change; it should not be presented as an objective absolute. Scores from different assessors are not directly comparable unless the method is applied identically. Mobile and desktop should be scored separately, as should Arabic and English.
Accessibility limitations. The criteria referenced are individual WCAG success criteria. Meeting them does not constitute conformance, and full assessment requires the complete standard.
Where review is required. Changes to what is collected at booking, and when, require privacy review under the Saudi Personal Data Protection Law, alongside clinical and administrative sign-off. This article is not legal advice.
Planned research. DEMA has scoped a Saudi booking friction benchmark covering a sample of healthcare booking journeys. Until it is completed and published, no benchmark figures exist and none should be inferred from this article.
Sources.
10
How hard is it to become your patient?
Most organizations have never scored their own booking path, and almost nobody has scored the Arabic one separately. A Healthcare Growth Diagnosis runs the index across every path a patient can take, separates the avoidable friction from the necessary, and prices the difference.
Request a Healthcare Growth Diagnosis