Larkspur Bank · 2025
Card controls people can find in a panic
Freeze, spending limits and merchant blocks for a mobile bank, designed for the anxious minute after someone realises their card is gone.
- Client
- Larkspur Bank
- Year
- 2025
- Role
- Senior product designer
- Duration
- 4 months
- Services
- Mobile design
- UX writing
- Usability testing
The problem
Larkspur is a mobile-only bank with about 600,000 customers. Card controls existed, but they were spread across three settings screens. When a card went missing, 63% of customers called support instead of freezing it in the app, and the average call waited 11 minutes. Fraud losses in that window were the bank's largest avoidable cost, and support was hiring just to keep up with lost-card calls.
I reviewed 200 recorded support calls with the fraud team and ran a first-click test with 120 customers on the existing app. Only 31% found the freeze option within 10 seconds. Interviews with 9 customers who had recently lost a card added the human detail: people were scared, often on the street, frequently using one hand, and they did not trust that a freeze was reversible. They wanted certainty, not options.
Process
- I
Put freeze on the card itself
I moved freeze to a single control on the card at the top of the home screen, reachable with one thumb, and grouped every other control on one card details screen. The risk was accidental freezes, so instead of a confirmation dialog I designed an unmistakable frozen state with one-tap unfreeze and a push notification. An accidental freeze costs a customer two seconds; a missed one can cost hundreds.
- II
Wrote every string with compliance
Copy carried most of the trust problem. I worked with the compliance and fraud teams on every string and tested three versions of the frozen-state message. 'Frozen. New payments won't go through. Unfreeze any time.' beat the legally cautious version in comprehension tests, 92% against 58%, and compliance approved it once we linked to what happens with pending transactions.
- III
Plain sliders and plain lists
Spending limits became a slider with presets based on each customer's typical monthly spend, and merchant category blocks became a checklist with plain names like 'Gambling' and 'Online shopping abroad'. We cut a requested feature, per-merchant blocking, because card network support for it was unreliable, and a control that sometimes fails is worse than no control at all.
Selected images
The lost-card queue used to be the first thing I checked every morning. Now I barely look at it.
Outcome
After launch, 71% of lost-card events were handled in the app without a support call, up from 37%. Median time from opening the app to a frozen card dropped to 6 seconds. Lost-card call volume fell 44% in the first quarter, which let the support team move four people to disputes. App store reviews that mention freezing went from mostly complaints to mostly relief.
- 71%
- lost cards handled in-app
- 6 sec
- median time to freeze
- −44%
- lost-card support calls
Keep reading
All work