Tools
I got my first real job as a designer at a startup called Nimbus working on a mental health app called Healu. Honestly, I had no idea what I was doing. I'd finished a design bootcamp three months prior, knew Figma basics, and had never done a real user interview in my life. But they hired me anyway—probably because I was cheap and enthusiastic.
The app was supposed to help people in the Middle East and North Africa (MENA) get mental health support. The region has a huge problem: lots of people with anxiety and depression, almost no therapists, and huge cultural stigma around therapy. The idea was to build an app with three things: a free AI chatbot (so people could get support anytime), paid therapy sessions with real therapists, and a community where people could share stories.
We spent 18 months building a prototype. I learned a ton. The app shipped to TestFlight beta with real users, but it had bugs and the team ran into problems with the engineering. It never made it to the App Store. But looking back, that doesn't matter as much as I thought it would. This case study is about what I actually learned by doing the work.
When I started, I didn't really understand what the problem was. The product manager told me: "People in MENA can't get therapy. Let's build an app." Cool. But I didn't get why that mattered or what the actual blockers were.
So I did interviews. This was terrifying. I'd never interviewed a real person about a product. I wrote down some questions, scheduled video calls via Zoom, and talked to people in Egypt, UAE, and Saudi Arabia. I remember one conversation with a woman in Cairo who said: "I've been anxious for years. There's one therapist in my city, but she's booked for six months. I can't afford someone private. So I just live with it." That stuck with me.
After talking to maybe 47 people (my manager suggested we keep a count), patterns emerged. People didn't use therapy apps because:
I made a quick Google Form and sent it to ~340 people. Same themes came up. I was starting to understand: this wasn't just about building an app. It was about trust and culture.
I made three different designs. Not because I'm some genius designer—I made them because I was confused and didn't know which direction was right.
Version 1: "Clinical Blue" — I thought healthcare should look clinical, like a hospital. Blue colors, serious fonts, lots of credentials shown. I literally copied some design patterns from hospital websites. It looked professional but kind of... cold? When I showed it to some of my friends, one said "this looks like where you go when you're really sick." That wasn't the vibe we wanted.
Version 2: "Cute and Friendly" — I swung totally the other way. Pastel colors, cartoon illustrations, made it feel like a wellness app. Users loved it. But then the therapist advisor looked at it and said "this doesn't look like a real healthcare product." She had a point. If I was looking for therapy, would I trust an app that looks like a children's game?
Version 3: The Middle Ground — I combined both. Warm colors (not clinical blue, but not pastel either), some illustration but professional too. I did like 20 rounds of tweaking the purple. This one tested better with both users and therapists.
I didn't have a fancy design system framework or anything. I just... tried stuff, showed it to people, and adjusted based on feedback. That's how I learned.
Here's where something clicked for me. Early on, the therapist profile in my design was super simple: name, photo, price. Three lines of text. Like Uber for therapy.
When I watched someone use it in testing, they spent like five minutes reading one therapist's profile—but there wasn't that much to read. They were looking for more information. They wanted to know: Who is this person? Can I trust them? What have other people said about them?
So I redesigned it. Now it had: a longer bio explaining who they are, what they specialize in, patient reviews (anonymized), and a button to message them before booking. Just... more context.
I watched conversion go up. Not by a crazy amount, but noticeably. People were more likely to book. That taught me something fundamental: people need to trust before they commit. Especially for something vulnerable like therapy.
This isn't rocket science, but it was my first time feeling the impact of a design decision through actual user testing. It made the work feel real.
By month 6, I had a full Figma file. Not a fancy design system with components and tokens—I didn't know what those were yet. Just a Figma file with like 50 screens that showed the whole flow:
The engineers looked at my Figma and built it in Flutter. I'd take screenshots, put them in Figma comments, say things like "this button should be bigger," and they'd adjust. It wasn't scientific. We just... communicated in Figma comments.
The team shipped it to TestFlight (Apple's beta testing app) after about 6 months. Real people could download it and actually use it. That was wild to me. Something I designed, other people could open and click around in.
We got 50 people to test it. They found bugs immediately. The payment button didn't work. The therapist calendar showed therapists as available when they weren't. The app crashed sometimes. Users couldn't figure out how to delete appointments.
These were things we should have caught, but we didn't. Because it was a small team and I was new, nobody had really done proper QA testing. We just... shipped it and hoped.
What was helpful: I got feedback from actual humans. One person said "I don't understand how to book an appointment. The button isn't clear." Simple feedback that made me realize the design wasn't obvious to people outside our team.
Another person said: "I love the mood tracking. I check in every day." That felt good.
Most people said the app had potential but "wasn't ready" and "has too many bugs." Fair.
By month 12, we had a working prototype. But there were problems:
By the end of 2024 (month 12), the product manager said "let's pause this." Not "this is dead," just "let's pause." We shipped something to TestFlight, got feedback, learned a ton. But the app never went to the App Store.
At the time, I was disappointed. I thought: "I failed. The app didn't ship." But looking back, that's not really what happened. We built something real, tested it with users, learned a lot. The fact that it didn't launch isn't the interesting part of the story.
Talking to users changes everything. Before I interviewed people, I had no idea why the app mattered. After, I understood the problem was real. People are actually stuck without therapy options. That context made every design decision feel meaningful. I'm never designing anything again without talking to at least some real people.
Design is about trust, not just pretty colors. I spent a lot of time picking the perfect purple. But the stuff that actually mattered was: does the therapist profile feel honest? Can I see what other people think of this therapist? Can I message them before committing? Those things moved the needle more than any color choice.
Shipping to real people is worth 1000 user testing sessions. TestFlight was messy and found problems. But those problems were real—not hypothetical. Someone couldn't complete a payment. Another person didn't know how to book. That feedback was worth more than any feedback from colleagues.
I don't know what I don't know. I didn't know about design systems, I didn't know about accessibility, I didn't know about how therapist workflows actually work. I learned some of these things during the project. Others I still don't fully understand. That's okay.
Shipping something broken is better than shipping nothing perfect. We shipped to TestFlight with bugs. It would have been "better" to ship with zero bugs, but we never would have shipped. The imperfect version taught us more.
What actually worked:
What didn't work:
We never launched to the App Store, so I don't have "millions of downloads" to show. But here's what I'd estimate if we had shipped:
We had 8% of TestFlight users actually book a therapy session. If that had scaled, we'd have had maybe 10K downloads in the first year if we'd been aggressive with marketing. Not a venture-scale success, but a real product with real users.
Ship faster. I spent way too long perfecting early designs. I should have shipped a rough prototype to users in month 2, not month 6. Feedback would have been faster and I could have iterated quicker.
Involve the therapists way earlier. We thought about the therapist side last. That was backwards. They're half the product. I should have designed their experience in parallel with user experience.
Test payment and booking flow with real users earlier. These were the critical flows, but we didn't test them thoroughly until TestFlight. By then it was too late to fix big problems.
Don't overthink the design. I spent weeks deciding on colors, spacing, typography. Most of that didn't matter. What mattered was: does the therapist profile feel trustworthy? Can someone actually book? Does the AI chatbot feel helpful? Focus there.
Be more honest about what we don't know. We assumed the business model would work (30% conversion). We should have tested that assumption with users earlier instead of discovering month 12 that it didn't.
The biggest thing I learned isn't about design. It's that shipping something imperfect teaches you more than perfecting something unshipped. TestFlight was a mess. Users found bugs. The conversion was lower than expected. But those learnings happened because something real existed. If we'd kept iterating in Figma forever, we'd never have discovered the actual problems.
I also learned that most junior designers overthink everything. I spent so much time worrying: "Is this the right color?" "Should this button be bigger?" "Are the margins perfect?" Meanwhile, the stuff that actually mattered (therapist trust, clear booking flow, simple payment) wasn't fancy design. It was honest design.
If I could tell a junior designer version of myself one thing, it's: stop perfecting and start shipping. Get real feedback. Iterate based on that. The feedback from one real person beats your own intuition every time.
The project is paused. The team moved on to other things. The Healu prototype sits in Figma, archived.
But I didn't stop learning. Every project since then, I've remembered:
If Healu taught me anything, it's that shipping something real and learning from it beats designing something perfect in Figma forever. The app didn't make it to the App Store, but the lessons did.
This wasn't a success story. There's no "we launched and got millions of users." But it's a real story. I was a junior designer who didn't know much, I got thrown into a real project, I learned by doing, I shipped something to real people, and it taught me more than any course or tutorial ever could.
If you're a junior designer reading this and thinking "I don't know enough yet," that's the point. You learn by doing. Ship something imperfect. Get feedback from real people. Iterate. That's it. That's how you actually become a designer.
