Two friends, one bill.
Bank built for both.
Shared accounts, four split modes, and a payment surface designed before it was built. MyBank asked for collaboration features. We asked the users first.
Three pain points the platform could not see.
Surveys, interviews, account-data review. The friction was structural, not surface, and three patterns kept appearing.
Banking is single-account by default.
Households and small teams share money in practice but track it in spreadsheets, screenshots, and photographs of receipts.
Visibility ends at the transfer.
Once money moves, the trail breaks. Who paid what, when, and on whose behalf becomes a follow-up message, not a ledger row.
Splits are arithmetic, not finance.
Even-split, custom-split, percentage-split, item-split: every group has a default the bank does not know.
One transaction. Four ways to settle it.
We mapped the four splits users actually used and built the interface around the choice, not the calculation. Tap a mode to preview the share.
Dinner with three friends. Equal share, no maths.
Designed on paper. Tested on people. Built by engineers.
- 01
Surveys + interviews
Quantify the gap, hear the friction in users' own language.
- 02
User flow design
Single information architecture for shared and solo accounts.
- 03
Usability testing
Two rounds, before any production code, before any visual design.
- 04
Build, with engineering
Pair-programmed UX; iteration on real data, not screenshots.
- 05
Handoff + launch
Done means used. Pushed live with the bank's release calendar.
Five surfaces. One mental model.
Balance overview
Solo + shared, single screen.
Quick actions
Transfers, requests, reminders.
Even split
Default for groups.
Custom split
By amount or percentage.
Settlement view
Who owes whom, at any point.
A complete toolkit for shared finances. Designed before it was built. Used the day it shipped.
