Case study · Mobile banking · via BCG

The bank let us redesign one flow. It grew into fourteen.

A year redesigning a major Indian bank’s mobile app with BCG: payments, login, and the design system behind them.

Role
Product Designer
Team
5 designers
Duration
12 months
Scope
14 core journeys
Plus
Design system
Entry 01The problem

The old app worked, but it was hard to use.

Navigation was confusing, key details were missing when people needed them, and the flows didn’t match Google Pay or Paytm, the apps customers already used every day. The bank had no brief, just a feeling the app wasn’t good enough.

Exhibit AThe existing flow
Source account screen
01Source account
Destination screen
02Destination
Enter amount screen
03Enter amount
Confirm details screen
04Confirm details
Verify with OTP screen
05Verify with OTP
Success screen
06Success
The old add-money flow. It worked, but the order of steps felt unfamiliar.
Entry 02The test flow

Before giving us the whole app, the bank tested us on one flow: adding money.

It’s used daily, so it was a fair test. Instead of new patterns, I used the ones people already trust from Google Pay and Paytm. When moving money, familiar feels safe.

Where I pushed back
The bank wanted to advertise its savings rate inside the transfer flow.
Problem

An ad in the middle of a payment distracts people when they most need to focus.

What I did

Removing it wasn’t an option, so I argued to make it smaller and keep it out of the main task.

Result

It shipped small and quiet. The first time the client went with my reasoning over their own.

Exhibit BThe redesign
Select account screen
01Select account
Enter amount screen
02Enter amount
Confirm details screen
03Confirm details
Enter UPI PIN screen
04Enter UPI PIN
05Success
Confirmation screen
06Confirmation
The new flow: pick an account, enter an amount, confirm, enter UPI PIN, done. The savings promo is there, but stays quiet.

The test worked. We got thirteen more journeys, including onboarding, payments, savings, credit cards, mutual funds and spend tracking.

Entry 03Login

People were getting locked out, so we rebuilt login.

Everyone sees the login screen, which makes it the easiest place to lose people.

Login redesign
One screen, one way in. If your details didn’t match, you were stuck.
Problem

A forgotten username, a new phone or a new SIM locked people out of their own account.

What we did

Four ways to verify: Customer ID, PAN, account number or debit card. SIM checks run in the background.

Result

Set an MPIN once, then log in with a fingerprint every time after.

Exhibit CBefore, the old login
The original IDFC FIRST login screen
Before: one screen, two fields, no backup if they didn’t work.
Exhibit DAfter, the full onboarding overhaul
Mobile number screen
01Mobile number
Verify SIM screen
02Verify SIM
Customer ID screen
03Customer ID
Create MPIN screen
04Create MPIN
Biometric screen
05Biometric
Welcome screen
06Welcome
After: mobile number, SIM check, ID, MPIN, fingerprint, done.
Entry 04Design system

Five designers made ten versions of the same button, so we built a design system.

We locked shared components and wrote down the patterns. Builds and reviews got faster.

It also helped with the client. They often asked for “something different”. The system gave us a clear reason to say no, and a quick way to say yes when a change made sense.

Exhibit EThe system
Component library, the design system
The component library every later flow was built from.
OverviewThe redesign, at a glance
The redesigned IDFC FIRST app, onboarding through dashboard
The full redesign, from login to dashboard.
Entry 05Results

No launch numbers came back to us, but the project kept growing.

We handed off to the bank and BCG before launch data was available.

Entry 06Learnings

Earn trust by delivering, and build the system early.

Trust comes from delivering. Good work earned me the right to push back, and taught me to pick which fights matter.

A design system is a tool, not a chore. It sped up the team and gave us a clear way to handle client requests.