Financial Management Tool

The constraint that came before any design work.

The tool opens inside the Capital One app, from the account home screen with a swipe, underneath the checking balance, credit card and credit score. That was decided before I started, and once you take it seriously it settles a surprising number of arguments.

What it gives you is enormous: trust is already there, transaction data is already present, and there is no signup. What it costs is real too: there is no tab bar of its own, no app icon, and every pattern has to match the parent app.

Less work became the product principle.

I started with requirements analysis alongside business stakeholders, then talked to people who already used budgeting tools. I had assumed people wanted more control over their money. What they actually said was that they wanted less work. That correction reshaped the feature list.

Low-fidelity flows came first, mapped around the paths a person takes: set a goal, build a budget, track a bill, read the trend. Then came high-fidelity screens, an interactive prototype, usability sessions, A/B comparisons on calls to action and navigation, and a build with engineering aligned to iOS Human Interface Guidelines.

A feature list is a wish list until every feature has a goal.

The findings became twenty-nine candidate features. Every one got tagged with the user goal it served, then sorted into four tiers with stakeholders in the room rather than by me alone. The tagging made the sort a question anyone could check: which goals will ship, and which remain only ideas?

Advice is a promise. Accuracy has to come first.

Eight features in the backlog would have given people financial advice: educational tips, hypothetical scenarios, suggestions about where to save, reports of unusual spending, a virtual advisor, a financial coach, and service integrations. Not one was must-have. One was nice-to-have, five were surprising and delightful, and two were parked.

The easy version of this product ships the virtual advisor early. It demos beautifully and is what every stakeholder gets excited about. We did not build it. A budgeting tool has to be boringly accurate about what already happened before it earns the right to have an opinion about what should happen next.

Reading the backlog by category turns that first finding into something more specific than a feature list.

Twenty-nine features do not become twenty-nine places.

They collapse into the four questions a person asks about their money: what am I saving toward, am I spending more than I meant to, what is about to come out, and where is the money going over time.

Goals, budgets, bills and cash flow became the reachable structure. Filters, custom categories, overspending alerts and scheduled reports happen inside those four. Anything that could not find a home had to justify becoming a fifth, and nothing did.

Money arrives with emotion already attached.

People arrive at a budgeting screen already braced for bad news. A screen full of accurate figures with no interpretation makes that worse rather than better, so every summary answers the question before it shows the arithmetic.

The goals card says two current goals and “no actions needed.” The budgets card says twenty-one budgets and “all on track.” The budget ring says $1,945 and “is looking good.” The credit score card already used the same pattern with 817 and “excellent.”

A reassuring interface that is wrong is worse than a neutral one. Reassurance is a small promise. Advice is a large one, and you should not make the large one until you have kept the small one for a while.

What the design work produced.

• End-to-end ownership of one product, from an unsorted feature list to a tested interactive prototype
• A prioritized backlog of twenty-nine features across four tiers, each tagged with the user goal it serves
• User flows covering the four core paths, plus high-fidelity screens and a working Figma prototype
• A component library unifying the three core flows, which engineering built from
• Usability findings and A/B comparison results on calls to action and navigation, folded back into the design
• Accessibility guardrails on contrast, tap targets and error prevention

A product team effort, with clear ownership.

A product team effort, with business stakeholders, research participants and engineering shaping what got built; my ownership was the research synthesis, the flows, the high-fidelity screens and prototype, the component library, the usability and comparison testing, and the interaction and accessibility specifications. I ran the sessions and comparisons but no longer hold the data, and this is the oldest work in the portfolio: the thinking holds up; the visual craft is of its time.