Building for People at Their Most Vulnerable: What Healthcare Taught Me About Mobile Product

Most product work happens when people are basically fine.

Someone opens a shopping app because they want new sneakers. They open their banking app to stare at a number they already know and feel a feeling about it. The stakes are low, the mood is neutral, and if your app annoys them, they sigh, toss the phone on the couch, and forgive you by morning.

Health apps do not get that grace.

When someone opens a health app, something is usually wrong, or they are scared something might be. They are exhausted, or in pain, or sitting in a parking lot doing the very American math of whether what they are feeling is worth a copay and half a vacation day. I spent years building in that world, across Fortune 5 healthcare platforms and a care accelerator, and I will tell you up front: no amount of SAFe training prepares you for the moment you realize a confused tap on a screen might mean a person does not get care. That is a different kind of stakes than a cart abandonment dashboard.

Here is what it taught me.

Your user is not at their best, and your UI has to assume it

In every product review, we secretly picture a calm, rested, well caffeinated person tapping serenely through our flow. That person is a myth. She does not exist, and she definitely does not exist in healthcare.

The real user is one handed, because the other arm is holding a feverish toddler. She is on a basement connection with two bars and a spinner that will not quit. She is reading your screen through tears, or through the specific brain fog of being genuinely unwell. Every extra field, every clever gesture, every place where she has to stop and decode what you meant is a place where a frightened person gives up and calls a 1-800 number. Or, worse, does nothing at all.

That one assumption reorders your entire backlog. Thumb reach matters. One handed flows matter. Cognitive load stops being a line item and becomes the whole game. When we built the See a Doctor pillar inside Evernorth’s Future of Care work, the wins that mattered were never the flashy ones. They were the moments we deleted a decision a scared person should never have had to make in the first place. The best healthcare design often looks like nothing happened. The friction simply was not there. Nobody tweets about the friction that was not there, which is a shame, because it is the hardest thing we do.

Trust is the actual product

People will hand a retail app their credit card without blinking. They will not hand a health app their symptoms, their history, and their fear unless they truly believe it will be held with care.

And nothing torches trust faster than an app that acts like it knows too much. You know the feeling. You mention something out loud once and suddenly your phone is suggesting it, and a little chill runs down your spine. In retail that is mildly creepy. In healthcare it is a breakup. One badly timed push notification, one permissions dialog that asks for everything at launch like a stranger demanding your house keys, and the user quietly pulls back. She gives you less. She stops telling you the truth. And a health product running on half truths is worse than useless, because it makes confident decisions on garbage data and feels great about it.

So I learned to treat trust as a feature with its own roadmap. How we ask. Why we ask. What we visibly choose not to do with what we collect. Permission requests that arrive in context, when the value is obvious, instead of a wall of dialogs in the first ten seconds. That is product work, not a legal footnote, and it deserves the same care as anything on the screen.

Clarity beats cleverness, every single time

In consumer apps, a little ambiguity can feel charming and on brand. In healthcare it is dangerous.

If a user is not certain whether she just booked that appointment, that is not a minor UX gap. That is a person who might miss care, or double book, or sit at home assuming help is coming when it is not. Confirmation states, plain language, and brutally honest status are not polish. They are safety equipment.

I got allergic to cute copy in this world. The goal was never to sound clever. The goal was for a frightened person to read one sentence, understand exactly what is true and what happens next, and feel her shoulders drop half an inch. If your microcopy makes a scared person calmer, you nailed it. If it makes them reach for the dictionary, you failed, and you failed in the one domain where failing is not funny.

The handoff is where care actually breaks

Here is the part nobody warns you about. The hardest part of a health product is rarely the part on the glass. It is the seam between the beautiful mobile experience and a real human being on the other end.

A gorgeous intake flow that dumps a patient into a dead end, or hands a confused human on the back end zero context, erases every good thing that came before it. The app made a promise. The system has to keep it. So I spent as much time on what happens after the tap as on the tap itself. Who picks this up. What do they see. How fast. Does the person feel cared for, or does she feel like a ticket.

Your analytics will lie to you here if you let them. The funnel looks gorgeous right up until the moment the human handoff drops the ball, and the only place you see it is in the support queue and the one star review that just says, in lowercase, crashes when i need it. That review is not about a crash. It is about a promise.

Why this travels way beyond healthcare

You might never build for sick people. But every product has users who show up scared, rushed, or in over their heads. The new hire on day one. The parent at 2am. The person who just got bad news and needs your tool to not make it worse.

Healthcare taught me to design for that person first, not last. Assume reduced capacity. Earn trust instead of demanding it in a permissions dialog. Choose clarity over charm. Sweat the handoff like your reputation depends on it, because it does. Do that, and your app becomes the kind people reach for in the moments that actually matter.

That is the only kind of indispensable worth building. Everything else is just another icon waiting to get swiped into a folder called Maybe Later.