Personal Risk Score
A five-minute questionnaire turned into a risk score you can act on.
Mobile
Fintech
Concept
Role
Sole Product Designer
Phase
Design Concept
Scope
48 screens, end-to-end
Platform
Mobile · iOS & Android
Team
1 designer
Methods
User Flows · Interaction Design · Visual Identity · Accessibility
Two ends of the app: the login screen that sets the tone, and the score it leads to.
Overview
A two-page brief, a scoring formula, and no product yet.
The client, a European startup, came with a clear mechanism: ask people about their life — work, income, family, worries — and turn the answers into a personal risk score. A premium tier splits the score into income, family and health risk, with recommendations to bring each one down. My job was to design the whole app around that mechanism: 48 screens covering login, a branching questionnaire, the score, the premium detail view and settings.
Problem
The app asks about money, family and fear of war — and the reward comes last.
Every step of the questionnaire touches something sensitive: income, dependents, worries like unemployment, healthcare or war. The user pays that price up front and only sees the value — their score — at the very end. A long form here would lose people halfway.
The output carried its own risk. The brief defined scoring as a traffic light per category, and a red light with no explanation is the kind of result that makes people close an app.
The pause, then the payoff: one number, plain words, a person instead of a gauge.
Key decisions
The main design decisions, and what they cost.
Decision 01 · Questionnaire
One question per screen, one tap per answer
The questionnaire never shows a form. Each question gets its own screen, and each answer is a full-width row you can hit with a thumb. The selected one turns blue. Questions that ask for a count get a row of tiles instead, and the multi-select question about concerns gets checkboxes. Skip stays available throughout, a hard requirement from the brief.
Occupation changes what comes next: employed, self-employed, unemployed and retired each get their own income ranges and their own list of concerns. That is how a handful of questions grew into 37 questionnaire screens. And where the brief listed two gender options, the design ships with three, each with its own hand-drawn bust.
Why
Sensitive questions are easier one at a time. Full-width answers also hide the branching: the user never sees the logic, just the next simple choice.
Trade-off
More taps, and far more screens to maintain. A single form would be quicker to build, but it would put income, dependents and your biggest fears into one intimidating view.
Decision 02 · The score
The traffic light became a ranked list that explains itself
The brief asked for a literal three-color traffic light per risk area. The design keeps the three colors but gives each risk its own shape and ranks them by size. Income, the highest risk, gets the largest shape and the top row. Health, which is stable, gets the smallest. Each row carries a status label and a sentence saying what it means and what to do about it. The home screen repeats the three shapes under the score, so the ranking is visible before you open the details.
Tapping a row opens the recommendations from the brief: concrete actions with checkboxes and a progress bar, so lowering a risk works like a to-do list instead of a warning.
Why
Color alone can't carry the message (WCAG 1.4.1) — roughly 1 in 12 men can't rely on red vs. green. A red light without words alarms without informing.
Trade-off
Less literal than what the client sketched. The at-a-glance dashboard effect is softer: you read a ranked list instead of scanning three dots. And size as a signal has a cost of its own. The stable risk is small on purpose, which makes it easy to overlook.
Decision 03 · Identity
Ink, one electric blue, and a hand-drawn cast
The interface is black ink on off-white, with a single blue: #0028F9. It floods two screens, the login and the loading moment before the score. On every other screen it marks the thing to tap: the selected answer, primary buttons, links. Risks get flat shapes in orange, yellow and green, so a status reads at a glance without a dashboard.
Everything human in the app is drawn in black ink: the group on the login screen, the busts on the gender question, the woman who delivers your score. The cast is deliberately mixed, a wheelchair, a prosthetic leg, different ages, because the app's premise is that risk applies to everyone.
Why
The subject matter is heavy, and a sterile fintech dashboard would make it heavier. Illustration gives the app warmth without pretending the topic is fun.
Trade-off
A strong illustration style dates faster than a neutral system, and future brand work has to reckon with it. The layouts hold illustrations in their own slots, so swapping styles later stays cheap.
Outcome
A buildable concept, before any code.
48 screens
the full journey — login, questionnaire, score, recommendations, settings
4 paths
occupation branches, each with its own income ranges and concerns
3 risk areas
income, family, health — each explained in plain words
The deliverable was a clickable Figma prototype of the full journey: 48 screens, four occupation paths mapped end to end, and the branching and thresholds from the brief carried into the screens. It was a pitch, so there are no usage numbers. What it settled was scope: the mechanism in the brief fits a five-minute questionnaire and a single results screen without dropping a rule.
The two moments the concept was built around: the wait before the score, and the checklist that lowers it.
Reflection
A rule-based spec doesn't have to look like one.
The brief was essentially a decision table: branching questions, threshold scoring, a paid tier. The design work was translation, keeping every rule intact while making each screen feel like a conversation. Phase II adds AI-generated recommendations, and the structure is ready for it: the list and the checklists don't care whether a human or a model wrote the advice.