Skip to the case study
Back to work
Jupiter app icon

Redesigning onboarding to add $2.4M in deposits

Jupiter.Money is an 🇮🇳 Indian fintech and digital banking platform serving over 3 million customers, with a valuation of $600 million

My role

End to end product design

Team

1PD, 2PMs, 6 Engineers + more

Duration

1.5 months

What did I do?

  • End-to-end ownership
  • User research
  • Impact modelling
  • Stakeholder alignment
  • Journey mapping
  • Competitive benchmarking
  • Interaction design
  • UX writing
  • Usability testing
Splash screen: an orbiting icon system around the Jupiter mark, with the line Your financial wellness account Process map screen: three completed onboarding steps marked with green ticks on a connecting spine, with account activation up next Product selection screen: a single featured recommendation, Open an all-in-1 Savings Account, with its benefits, imagery and a social-proof stat, followed by an Explore more on Jupiter row of additional product cards Selfie check screen: the word selfie replaced with take a picture to verify it's you, an illustrated example, and three guidance tiles for lighting, framing and no glasses or caps

Prologue

Jupiter had evolved from a wallet to a full stack finance app

Earlier

A space-themed savings account, largely used as an expense wallet

Audience

Students and youngsters, tracking daily spends

Jupiter's original splash screen: the Jupiter wordmark over the line 1-app for everything money, an astronaut planting a Jupiter flag, and the greeting Look who just got a window seat to Jupiter.
Jupiter today: a savings account showing a balance of 56,000 rupees with bank transfers, fixed deposit and debit card, over rows for loans, credit cards and investments.
Now

Added an entire line of banking products like Loans, Credit Cards, Investments

Audience

Financially mature with higher disposable incomes

The ask

Optimise the onboarding funnel to improve conversion

Since the uber goal was improving M1 retention, and my hunch was that this alone would not be enough. High Savings Account rejection rates and low M1 activation were indicators of a deeper problem. So I ran a few quick studies to test my hypothesis.

  • ICP interviews

    "I thought Jupiter was more like a wallet for tracking spends"

    ICP match, never used Jupiter
  • Cold calls

    "I wasn't really sure what I was signing up for."

    Recent onboarder
  • Usability tests

    "Why do you need my PAN and Aadhaar? I just wanted to see my expenses."

    Live walkthrough

From funnel optimisation to a company wide big bet

The evidence gave us a bigger opportunity

The initial research findings and our projected business impact downstream helped broaden the scope from conversion optimisation to the wider onboarding redesign. With product, engineering and marketing aligned, onboarding became a company-level priority for the quarter

We reshaped the journey around the user's mental model

High dropoff points Lower dropoff points

  1. Verify mobile and email
  2. Verify PAN and give consents
  3. Get a "ticket" to Jupiter
  4. Select a product to onboard
  5. Complete e-KYC by verifying Aadhaar
  6. Liveness check for identity verification
  7. Add personal details, occupation and a nominee
  8. Set an mPIN for your account
  9. Add money to account
  10. Set Debit Card PIN & order card
  11. Finish video KYC
  • Sign up
  • Verify your PAN
  • Pick a Jupiter experience
  • Set up your account
  • Activate your account

Behind the scenes

Understanding users everyday lives beyond Jupiter

Spectrums highlight the range of temporary or permanent challenges of a person

  • Spatial context
  • Technological context
  • Familiarity with financial apps
  • Cognitive ability
  • Language & Literacy
  • Financial capacity
  • Security, Trust, Privacy
  • Intrinsic motivation
  • Comprehension

Situations are temporary challenges that might affect a person

  • Environment & surroundings
  • Security and trust sensitive
  • Connectivity & bandwidth
  • Time sensitivity
  • Document readyness
  • Error encounters
  • Temperament
  • Availability
  • Downtimes
Suraj, 26 Software Engineer at TCS, Bengaluru
  • ₹13L a year, ₹80K take-home
  • Moved from Ahmedabad for work
  • Shares a rented flat with two flatmates

"Everyone around me seems to have it all figured out. Meanwhile, I don't know how to start saving or where to invest."

Financial goals

  • Spend smartly — covers wants and needs while still saving for aspirational goals
  • Spending awareness — wants to know where his money goes each month
  • Investment guidance — asks friends and parents where and how much to invest
  • App expectations — simple, reliable, trustworthy money tracking

Pain points

  • Earning three years, but feels he hasn't saved enough next to his peers
  • Confused and overwhelmed about where to save and how to invest
  • Can't tell which platform fits his financial goals best

Trigger to download

  • Discovery — sees an Instagram ad about spend tracking and saving
  • Decision — downloads the app over lunch, while working from home
Raj, 29 Senior Analyst at PWC, Mumbai
  • ₹20L a year, ₹1.2L take-home
  • Lives in Mumbai with his partner
  • Planning to get married soon

"My basic needs are covered, but I want to be financially ready for future expenses while maintaining a comfortable lifestyle."

Financial goals

  • Smart savings — plans ahead for milestones like his wedding and buying a car
  • Goal-based investing — has some investing experience, now wants it aligned to specific life goals
  • App expectations — one app to track savings, investments and expenses, with room to explore loans when needed

Pain points

  • Decent savings built up, but worried about growing financial responsibilities toward parents and family
  • Confused and overwhelmed about where to save and how to invest
  • Can't tell which platform fits his financial goals best

Trigger to download

  • Discovery — hears about the app from a colleague who uses Jupiter for investments and mini-loans
  • Decision — downloads it one evening after coming back from office
  • NiyoXProcess visibility
  • MonzoClutter-free selection
  • N26Intent-based personalisation
  • FiVisual priming
  • CoinbaseBenefit-first permissions
Cross-industry
  • Pediatric waiting roomsCalming design
  • NoomTrust via tone and social proof
  • Festival signageGlanceable comprehension

Three recurring problems gave us three principles to design around

  1. 01
    Problem

    Expectation mismatch

    The flow felt too long for what users thought Jupiter was

    Principle

    Prioritise
    transparency

  2. 02
    Problem

    No trust

    PAN and Aadhaar felt disproportionate to the value users understood they were getting

    Principle

    Establish
    credibility

  3. 03
    Problem

    Missed cross-sell

    Onboarding didn't help users discover Jupiter's broader product ecosystem

    Principle

    Provide guidance with
    cross-sell

Splash

Making the first impression count

Before

An unclear value proposition followed by an immediate ask to act, before users knew what Jupiter was

Splash screen before redesign
Splash screen after redesign
After

Starting with a warm welcome that seamlessly led users into the journey, removing an unnecessary tap

Mobile number

Setting the right expectations from the start

Before

Asking for permission without sharing the reason or benefit

Mobile number screen before redesign
Mobile number screen after redesign: a savings-rewards graph above a sheet asking for the number
After

Combining the first step with the value proposition, and making permissions feel useful to the user

Permissions

Making permissions feel purposeful

Before

Multiple permission asks felt overwhelming and alarming this early in the journey

Permissions screen before redesign
Permissions screen after redesign: the three permissions named and reasoned, above an illustration of the system dialogs
After

Transparency put users at ease, while visual priming showed them what came next

Loaders

Using waiting time to build confidence

Before

Random facts filled the wait, but gave users little reason to care

Loaders screen before redesign: "Verifying your PAN number" over a generic Jupiter-orbit animation, with an unrelated planet fact filling the wait
Loaders screen after redesign: the wait is spent scanning the PAN card the reader just entered, with a line of social proof pinned at the bottom
After

Contextual visuals and social proof turned the wait into a moment to build trust and awareness

Process Map

Making the journey visible from the start

This new addition brought more transparency to the onboarding, helping users understand what was ahead and track their progress

Process map after redesign: four named onboarding steps on one spine, the current step highlighted with a time estimate, each completed step turning into a green tick

Realistic time estimates, clear steps, upcoming document requirements, and progress celebrated

PAN

Addressing the concern behind PAN drop-offs

Before

30% drop-off as users worried sharing their PAN could impact their credit score

PAN screen before redesign
PAN screen after redesign: the credit score concern answered in the subhead, the consent checkboxes moved down by the CTA
After

Addressing the credit-score concern directly helped halve PAN verification drop-offs

Product selection

Helping users choose the right product, faster

Better user underwriting enabled Jupiter to make smarter, more relevant product recommendations, reducing the cognitive load of choosing between multiple products

  • Scale for 3-6 products
  • Dynamic state-handling
  • First fold optimisation
Before

Overloaded with text and numbers

Product selection screen before redesign
Product selection screen after redesign: a single featured recommendation, Open an all-in-1 Savings Account, with its benefits, imagery and a social-proof stat, followed by an Explore more on Jupiter row of additional product cards
After

Strong product recommendation with ease for browsing more products. Actionable headings, supporting visuals, social proof, and backend control over card order

Iterations

A product selection system built to evolve with Jupiter

Product selection had to balance Jupiter's evolving product strategy with users' need to make a confident choice. We explored three strategic hypotheses, spanning user control to system intelligence, to find the right model

Selfie check

Removing the self-consciousness from verification

Before

“Selfie” made users self-conscious and unsure how their photo would be used

Selfie check screen before redesign
Selfie check screen after redesign: the word selfie replaced with take a picture to verify it's you, an illustrated example, and three guidance tiles for lighting, framing and no glasses or caps
After

A clear purpose and guided instructions made the step feel simple and reassuring

Error handling

Turning a dead end into a next step

Before

Playful, quirky metaphors made an already frustrating rejection feel more frustrating

Error screen before redesign: a full-screen sorry message reading 'we'll let you know when a new seat is available', with no explanation of why or any alternative offered
Error screen after redesign. It opens on an apology — Your Savings Account could not be opened, this can happen if your application didn't meet Federal Bank's guidelines — with a Help link in the corner, then reorganises: the error moves up and shrinks to a caption and a sheet rises offering free zero-balance Pots instead, with a balance illustration, two benefits and the note that 25% of users reduced overspending with Pots.
After

Gracefully handling the rejection with a personalised alternative path forward

Overall impact
14% improvement in overall funnel conversion, from 54% to 68%
10K new accounts created monthly, at ₹0 additional CAC
$2.4M additional annualised deposits

How I'd approach this today

  1. 01
    Research Granola · NotebookLM

    Transcribe and synthesise interviews faster, surface patterns across conversations, and use AI to challenge my initial assumptions

  2. 02
    Explore + Prototype Claude · Figma MCP

    Go straight to high-fidelity, working prototypes and explore multiple visual directions in the time it once took to make one

  3. 03
    Motion Claude Code · Three.js

    Push motion much further to explore how transitions and interaction can shape the visual experience, not just decorate it

  4. 04
    Test + Iterate Claude

    Put real prototypes in front of users earlier, then use AI to speed up synthesis and iteration while keeping the judgement human

What came next

  1. 01 A new entry point

    We experimented with an alternative app entry via TPAP, letting users link their mobile number and start making payments from an existing bank account through Jupiter. The home screen then became an opportunity to discover and cross-sell other products

  2. 02 Still evolving

    We also tested a browse-first home screen for another user segment. Early numbers were weaker, so we stayed product-led while continuing to experiment with user-quality segmentation, relevant product routing, and faster load and processing times

Jupiter app icon

Thanks for watching

Next project Back to work