From Figma to the App Store

The short version

Problem: Most finance apps feel like spreadsheets in disguise, and many insist on bank syncs that privacy-minded people do not want.

What I did: Researched, designed and tested the app, then built it myself in SwiftUI and shipped it.

Result: A live App Store app, and a design that held up in moderated testing: participants completed all five tasks.

The problem

People want to manage their money but get put off by cluttered interfaces, mandatory bank syncs and zero flexibility. The question: how might we make a budgeting app that is clear, motivating and respectful of people’s data?

Who I talked to

I surveyed 23 people aged 20 to 65+ and followed up with interviews. From that came three personas: Kevin (23, student and freelancer), Jane (38, curbing impulse spending) and Doug (67, retired and tech-cautious). The big findings: most people track manually, privacy is non-negotiable, one person’s “Tech & Gear” is another’s “Emergency Chocolate”, and visual summaries beat long lists.

Decisions

Manual tracking first. No accounts. Privacy-conscious users should not be forced to hand over bank access. Tracket has no accounts and does not sync bank data.

Customisable everything. Colour-tagged custom categories, plus a Hide Balance toggle for nosy commutes.

One thumb, please. After the first workshop, I moved expense entry to the bottom of the screen for one-handed use, refined the bottom navigation and merged profile and settings.

Calm, not clinical. Soft shadows, rounded corners, dark mode and accessible contrast.

Testing

Moderated think-aloud sessions, in person and remote, across five tasks: onboarding, adding an expense, setting a budget, viewing insights and adjusting settings. Participants completed all five.

Then I built it

I liked it enough to build it. Tracket is written in SwiftUI in Xcode, with AI assistance, and it is live on the App Store. Building it myself meant every vague corner of the design showed up fast: empty states, edge cases and what happens when the data is messy.

What it looks like

Three image frames: the design next to the finished build; the personas; usability testing.

Reflection

Designing for Doug and Kevin at once pushed me beyond my own use case. Next: bill splitting, animated feedback and longer-term studies of how habits change.