Mobile is the GLP-1 intake. Design for it that way.
Most GLP-1 intake traffic in 2026 starts and ends on a phone. The patient saw an ad on Instagram or TikTok, tapped through to a landing page, and entered the intake without ever switching to a laptop. That is the dominant journey. The desktop experience is the secondary one.
Yet most telehealth intake forms were originally built for desktop and made responsive on the way to mobile. A responsive form is not a mobile-first form. The patterns that lift completion on a phone are specific to the device, and the operators who design for mobile-first see meaningful gains, often without sacrificing the desktop experience at all.
This post is the practical playbook for mobile-first GLP-1 intake: how to design the screen, the questions, the inputs, the upload steps, and the session continuity so the funnel finishes on the device the patient actually uses.
For the foundational intake design patterns this builds on, see Intake Forms That Convert: Questionnaires for Modern Telehealth Programs and Smart Branching in Intake Forms: Fewer Questions, Better Qualification.
Mobile-first is a different design problem
A few structural differences worth keeping front of mind.
| Difference | What it means for intake design |
|---|---|
| The viewport is narrow | One column, with each element earning its place |
| Input is the thumb | Tap targets sized for thumbs, not cursors |
| The keyboard takes half the screen | Form fields and submit buttons must stay above the keyboard, not under it |
| Network is variable | Interactions feel laggy when they would feel instant on desktop |
| The user is doing other things | The session can be paused at any moment |
| The patient may switch devices mid-flow | Session continuity matters more than on desktop |
| Battery and data costs | Heavy assets, autoplay video, and large uploads cost the patient something |
| Accessibility settings vary | Font size overrides, reduced motion, color contrast all shift the design |
Each of these is a reason mobile-first intake is its own discipline. None of them is a reason to apologize for designing differently for mobile.
The thumb zones and tap targets
The phone is held in one hand most of the time. The thumb reaches most of the screen comfortably; some areas need a second hand or a stretch.
Design implications:
- Primary action lives in the thumb-comfortable zone at the bottom-center of the screen. The "Continue" button belongs there.
- Tap targets are at least 44 by 44 points, with comfortable spacing between them. Smaller targets miss in real-world conditions (in motion, on a couch, with a case on the phone).
- Inactive areas get padding so a misfire is forgiving. A patient who taps near the wrong answer should not be pushed to it.
- No important controls in the top corners. The pull-down for notifications and the system status bar are competing for the same pixels.
- Important actions are not hidden behind a hamburger menu during intake. A patient looking for the back button should find it where they expect.
Small UX wins here compound across a 15 to 30 step form. The patient who never has to think about how to tap the next button is much more likely to finish.
Single question per screen vs. progressive disclosure
The mobile-first default for intake is single question per screen. The screen is short, the focus is total, momentum builds with each tap.
The single-question pattern works because:
- The patient knows exactly what is being asked
- The screen never has too much going on
- Each completion is a small win
- The flow feels fast even when the question count is similar
Progressive disclosure (a longer screen with multiple questions revealed as the patient answers) can work for closely related questions (e.g., date of birth plus state, or two screens worth of demographics combined). The rule of thumb: keep it to two or three closely related questions per screen, never more.
A few patterns that hold up:
- Each screen has a clear primary question. Supporting questions revealed only after the primary is answered.
- No more than one question requires typing per screen on mobile. Typing on a phone is friction.
- Progress is visible. A simple progress bar at the top, or a step count, keeps momentum honest.
- Back and forward exist and behave correctly. A patient hitting back should land on the screen they expect, with their answer preserved.
For the broader intake structure decisions, see Smart Branching in Intake Forms: Fewer Questions, Better Qualification and The Weight Loss Intake Form Scorecard: 12 UX Checks Before You Launch.
Native keyboards: pick the right one every time
The keyboard is the single biggest UX surface on a mobile intake form. Picking the right one is one of the highest-leverage design decisions.
| Input type | Right keyboard | Why |
|---|---|---|
| Phone number | Phone numeric keypad | No typing letters or symbols |
| Email keyboard with @ and .com | Faster, more accurate | |
| Numeric value (height, weight, age) | Number pad with decimal | No accidental letters |
| Date of birth | Date picker, not text field | No format ambiguity |
| Single choice from a short list | Tap target radio buttons | No typing at all |
| Single choice from a long list | Searchable dropdown | Avoids scroll fatigue |
| Multi-select | Tap target checkboxes | One tap each |
| Free text (open-ended) | Standard keyboard | Use sparingly |
| Address | Address autocomplete | Reduces typing dramatically |
| ZIP code | Numeric keypad | Five taps, done |
Native pickers and structured inputs beat free-text fields almost every time on mobile. The single biggest mobile intake mistake is asking a patient to type something the device could capture another way.
Photo and document upload on mobile
The upload step is often the largest single drop-off in a GLP-1 funnel, and mobile is where it breaks most often.
A few patterns that hold up.
Camera or library, in that order
A clear primary "Take a photo" action with a secondary "Choose from library" option is more conversion-friendly than a generic "Upload" button. The patient who is opening the camera does not have to switch contexts.
Orientation and crop guidance
A visible guide ("Hold your phone flat, framed like this") returns more usable photos and fewer redos. A redo is often a drop-off.
File size handling on the device
Resize and compress on the device before upload. A 12 MB photo over a cellular connection is a five-second perceived delay that breaks momentum.
Document upload alternatives
For lab uploads, document scans, or insurance cards, offer alternatives:
- "Take a photo of the form"
- "Upload a PDF"
- "Email it to us at this address"
- "Send it via SMS"
- "Skip for now and upload later"
A program that holds the patient inside the funnel even when the document is not handy converts dramatically more upload steps than one that blocks until the document appears.
Deferred upload as a real pattern
Letting the patient finish the intake and complete payment, then handle the upload from a magic-link sent to email or SMS, recovers a meaningful share of would-be drop-offs. The provider review pauses until the document arrives; the patient is already in the program.
For the broader funnel measurement framing, see Find the Leaks: How to Instrument a GLP-1 Sales Funnel and Fix the Step That's Actually Costing You.
Session continuity across devices
Mobile intake sessions get interrupted. The patient steps away to do something else. The phone rings. They switch to a laptop because typing is faster for one question. They come back the next morning.
A program that survives these moments converts dramatically more.
The patterns that work:
- Resume by magic link. Capture email or phone early in the intake. Send an immediate "Pick up where you left off" link.
- State preserved on the server, not just the device. A patient switching devices should find the same answers.
- Clear "Save and resume" affordance. A patient who is not ready to finish should be able to step away on their own terms.
- No re-validation of already-answered questions. A patient resuming the form should not be asked again what they already answered.
- Sensible expiry. Sessions persist for at least 7 to 14 days; shorter than that breaks too many real-world flows.
The mobile-first version of this is to never assume the patient will finish in one sitting on one device.
For the related abandonment-recovery layer, see Abandoned GLP-1 Checkouts: A Recovery Flow That Wins Back Drop-Offs Without Feeling Pushy.
Perceived performance is real performance
On mobile, perceived speed often matters more than actual speed. A few patterns to consider.
First paint and immediate response
The first screen renders fast and the first tap registers instantly. Anything slower trains the patient to expect lag.
Optimistic UI transitions
A "Next" tap advances the screen instantly while the server saves the answer in the background. If the save fails, the screen handles it gracefully.
Skeleton loaders, not spinners
When loading is unavoidable, a skeleton layout is friendlier than a spinning indicator. The patient sees something taking shape.
Pre-fetch the next likely screen
Once the patient is in the flow, pre-fetching the next screen avoids the second-long delay between answers.
Minimal third-party scripts during intake
Analytics, A/B testing, and marketing pixels are useful, but each one slows the page. A leaner intake page outperforms a heavier one, particularly on slower connections.
A noticeably fast intake is one of the cheapest conversion wins available. The work is engineering discipline more than design.
Reduce typing wherever possible
Typing on a phone is friction. Every question that can be a tap should be a tap.
Patterns that reduce typing:
- Default to dropdowns, radios, and tap targets instead of free text
- Use device autofill for common fields (name, email, address, payment)
- Use address autocomplete to capture full addresses with a few keystrokes
- Use OAuth or magic-link patterns for authentication when applicable
- Allow document upload by photo instead of typed-in values where possible
- Re-use prior answers (date of birth filled, state filled) across the session
When typing is unavoidable, make it forgiving:
- Real-time validation that fixes formatting (phone numbers, ZIP codes)
- Generous tolerance for capitalization and spacing
- Clear, friendly error messages that name the actual fix
- Keyboard "Next" buttons that advance the field, not dismiss the keyboard
Less typing equals more completion. This is one of the cleanest patterns in mobile intake design.
Accessibility patterns that double as conversion wins
Accessibility design and mobile-first conversion design point in the same direction. The patterns that help users with low vision, motor impairments, or cognitive load also help every other patient under real-world conditions.
A few patterns:
- Generous tap targets help users with motor impairments and patients walking through an airport
- High contrast and sufficient font size help users with low vision and patients reading on a sunlit screen
- Reduced-motion mode respects users with vestibular sensitivities and reduces battery usage
- Voice-over and screen-reader compatibility helps users with visual impairments and any patient using device accessibility features
- Clear focus states help keyboard navigation and forgive imprecise taps
- Honest error messages help every patient, not just users relying on assistive technology
- Language and reading-level discipline broadens the audience and reduces support load
For the related accessibility patterns at the portal layer, see Accessibility for Telehealth Portals: The UX Checks That Reduce Support Tickets and Drop-Off. For the multilingual layer, see Multilingual Telehealth Intake: Why Language Access Is Becoming a Growth Lever.
Mobile-specific completion benchmarks
The right benchmarks for mobile intake are slightly different from desktop.
| Metric | What a healthy mobile intake looks like |
|---|---|
| Step-level completion | 85 to 95 percent per step in the easy stretches |
| Drop-off concentration | Identifiable at 2 to 3 specific steps, not spread across the form |
| Overall completion (intake start to provider review) | 50 to 70 percent for high-intent traffic |
| Time to completion | 5 to 12 minutes depending on program complexity |
| Mobile vs. desktop completion gap | Under 10 percentage points; closer is better |
| Photo or document upload completion | 70 to 85 percent when properly designed |
| Session resume rate (when used) | 25 to 40 percent of abandoners return to the magic link |
The honest goal is closing the gap between mobile and desktop completion, not optimizing them independently. Most programs ship a mobile experience that converts 30 to 50 percent worse than desktop and accept it. Closing that gap is one of the largest available wins in the funnel.
For the broader testing approach, see How to A/B Test Intake Forms Without Breaking Clinical Ops and A/B Testing for Telehealth: What to Test on Landing Pages and Intake Flows.
Implementation checklist
Use this when designing or auditing a mobile-first intake.
Layout and tap targets
- Primary actions in the thumb zone
- Tap targets at least 44 by 44 points with spacing
- Important controls not in top corners or behind hamburger
- Visible progress indicator
- Back and forward behave correctly
Questions and inputs
- Single question per screen as the default
- No more than one typed question per screen
- Correct native keyboard for every input
- Date pickers for dates
- Address autocomplete for addresses
- Defaults to tap-based inputs over typing
Upload step
- Camera-first upload action
- Orientation and crop guidance
- On-device resize and compression
- Document upload alternatives (PDF, email, SMS, deferred)
- Deferred upload pattern with magic link
Session continuity
- Email or phone captured early
- Server-side state persistence
- Magic-link resume
- At least 7-14 day session expiry
- No re-validation of completed answers on resume
Performance
- First paint and first tap respond instantly
- Optimistic UI advancement
- Skeleton loaders for unavoidable waits
- Pre-fetch of next likely screen
- Minimal third-party scripts during intake
Accessibility
- High contrast and adjustable font size
- Reduced motion respected
- Screen-reader compatibility tested
- Clear focus states
- Plain-language error messages
Measurement
- Step-level completion tracked separately for mobile and desktop
- Upload completion tracked specifically
- Session resume rate tracked
- Time-to-completion tracked
Final takeaways
Mobile-first intake is one of the highest-leverage conversion improvements available to a GLP-1 telehealth brand in 2026.
What to remember:
- Mobile is the GLP-1 intake; design for it that way
- The viewport, the thumb, the keyboard, and the variable network all shape the design
- Single question per screen is the default, with progressive disclosure used sparingly
- Native keyboards and structured inputs beat free text almost every time
- The upload step is where mobile intake most often breaks; design specifically for it, including deferred upload
- Session continuity across devices and time recovers a real share of would-be drop-offs
- Perceived performance is real performance; engineering discipline pays back at the conversion level
- Less typing equals more completion
- Accessibility patterns and conversion patterns point in the same direction
- The honest goal is closing the mobile-to-desktop completion gap, not designing them separately
A brand that takes mobile-first intake seriously is leaning into the device the patient actually uses, in the moment they actually use it. The patients finish more often, the experience reads as more thoughtful, and the funnel does what it is supposed to do: get the right patient to the right care, on the device they were already holding.