Guide
Conversion on a regulated health form
Every growth playbook written for ecommerce assumes you can shorten the form, add urgency, and put a happy customer's face next to the buy button. On a prescription-only medicine, each of those is either unlawful or actively being enforced against right now. That is not a reason to accept a leaking funnel. It is a reason to pull different levers.
This page sets out both halves: the tactics that are off the table on a UK online pharmacy or prescriber flow, and the ones that are legitimate and still move the number.
What's off the table
Shortening the clinical questionnaire
The GPhC's April 2026 review of weight management inspections names inadequate risk assessment, and reliance on self-reported height and weight, as the sector's core failing. Trimming clinical questions to lift completion rate is the precise practice the regulator is currently enforcing against. Changing how the questions are presented is fine. Removing them is not.
Patient testimonials and outcome stories
Prescription-only medicines cannot be advertised to the public under Part 14 of the Human Medicines Regulations 2012. A testimonial describing what a named drug did for someone is advertising, not social proof. Service-level reviews that do not name or imply a prescription-only medicine are a different matter, and are the version to use.
Countdown timers, "only X left", limited-time framing
Falsely claiming a product will only be available for a limited time is on the automatically-unfair list under the Digital Markets, Competition and Consumers Act, in force since 6 April 2025. It also sits badly beside the GPhC's separate concern that commercial pressure should never make a prescriber hesitate to decline.
Pre-ticked marketing consent, or bundling marketing opt-in with clinical consent
The ICO and CMA's joint work on harmful online design names leading language and bundled choices that force data sharing. Health data is special category data, so the consent standard here is higher, not lower. Marketing consent must be a separate, unticked, genuinely optional choice from the consent needed to proceed with care.
Charging before the clinical outcome is known, or hiding the real price until late
Two risks stack. The GPhC does not want a financial arrangement that makes a prescriber hesitate to decline. And the DMCCA requires the total price, including any compulsory fees, to be shown prominently in the invitation to purchase rather than dripped in later.
What's legitimate, and likely to move the number
Progressive disclosure of the same question set
One question, or a small related group, per screen instead of one long scroll. Nothing is removed, only re-presented. This is the safe version of "shorten the form", and in practice it is the single biggest completion lever on a long clinical assessment.
Save and resume
Identity checks and photo uploads are natural abandonment points. Letting someone return to a part-completed assessment, rather than starting again, recovers those people without touching a word of clinical content.
Start before the signup wall
Let the assessment begin the moment someone lands, and create the account once there is something worth saving. Registration walls in front of the first question cost more than they protect.
Your GPhC registration number as the trust signal
A registration number in the footer, linked to the public register, beats any badge you could design, because the patient can click through and verify it themselves. The GPhC maintains a public, fully up-to-date register of pharmacists, technicians and premises.
"Why we're asking this" microcopy
Placed next to the harder questions, this reduces drop-off caused by confusion rather than by rigour. A patient who understands why a question matters answers it more carefully and abandons less often.
Charge for the consultation itself, disclosed in full upfront, regardless of outcome
This removes the "pay to find out whether you're declined" surprise that kills conversion at the worst point in the funnel. It also happens to be the structure the GPhC prefers, for the different reason that no fee is tied to an approval.
Equal-prominence cookie and consent choices
No pre-selected defaults favouring more data collection. Accept and reject given equal weight. This is the standard the ICO and CMA are enforcing elsewhere, and an obvious dark pattern on a health form costs more in credibility than it ever returns in data.
How Sync-RX supports this
- One question per screen with branching, so the full clinical set is asked without a wall of fields.
- Part-completed assessments are saved and resumable by secure link.
- Photo-on-scales capture and identity checks, so BMI is independently verified rather than taken on trust.
- Consent items are individually configurable, separately recorded, and never pre-ticked.
- Per-question help text for the questions patients stumble on.
- Full audit trail of every clinical decision, including refusals.
FAQ
Can we A/B test our questionnaire?
Yes, on presentation: wording, order, screen grouping, help text, imagery. Not on which clinical questions are asked. Sync-RX's A/B testing is built for exactly that split.
Can we use reviews at all?
Yes, for the service. Not for a prescription-only medicine or its effects. Keep review content on the pharmacy, the delivery and the support, never on the drug.
Is a discount code allowed?
Not on a prescription-only medicine. Promotional discounting of prescription-only medicines is the subject of joint ASA, MHRA and GPhC enforcement.
What counts as a compulsory fee for pricing rules?
Anything the patient cannot avoid paying. Where a fee depends on choices made later in the flow, the position is less clean, and the CMA's price transparency guidance is the document to work from with your own adviser.
Does any of this reduce conversion?
Clinics that move to progressive disclosure with save-and-resume usually see completion rise, not fall. The compliant levers and the effective levers overlap more than the ecommerce playbook suggests.
Book a walkthrough
Want to see a compliant assessment flow running end to end?