Skip to content
Consumer fintech · UPI payments Shipped

Tata Pay UPI · Tata Digital · Tata Neu

Making UPI payments feel as trustworthy inside a super app as in a bank's own app.

I led the payment experience for Tata Pay inside Tata Neu: reorganising it around what people intend to do, keeping linked-bank status always visible, and treating failed payments as moments to reassure rather than dead ends.

Trust & safety Payment states Mobile Design tokens
Tata Pay mobile screens across transaction lifecycle: onboarding, scan & pay, and instant confirmation.
Tata Pay mobile screens across transaction lifecycle: onboarding, scan & pay, and instant confirmation.

Executive snapshot

Product
Tata Pay UPI within the Tata Neu super app
My role
Senior UX Designer and FinTech design lead
Platforms
iOS, Android, responsive web and SDK
Payment rails
NPCI UPI · multi-bank · P2P and P2M
Partners
Product, risk, payments engineering, brand teams
Users
Indian consumers across Tata Neu brands
~42%Fewer payment steps
Reported outcomeSteps to complete a standard payment, before and after redesign. Baseline and measurement period being confirmed.
+22%Task completion
Reported outcomeSuccessful completion of core payment flows. Baseline and measurement period being confirmed.
+33%Self-serve recovery
Reported outcomeFailed payments resolved by users without contacting support. Baseline and measurement period being confirmed.

01 · The stakes

Trust was the product.

Tata Pay runs on India's UPI rails, where people expect near-instant confirmation and bank-grade reassurance. Inside a multi-brand super app, any hesitation in the payment flow reduced checkout conversion across every brand.

The real problem was not adding UPI. It was making money movement feel transparent and consistent when people juggle several linked bank accounts, mandates and frequent bank-side downtime.

02 · My leadership contribution

Lead the payment experience end to end and create the patterns other Tata Neu teams would reuse for checkout.

Team & collaboration model
Design lead for the payments stream, partnering closely with product, risk and payments engineering.
Key decision
Model every transaction state with engineering and risk, then design explicit, honest messaging for each one.
Stakeholder challenge
Aligned brand teams who wanted their own checkout look with a shared, trusted payment layer.
Operating rhythm
Weekly reviews with risk and engineering on payment states; shared pattern library for downstream teams.

03 · Research & insight

What we learned before we designed.

Users compared us to their bank

Any visual hesitation or unfamiliar wording made people doubt the payment had worked.

Failures were misread as crashes

Bank downtime looked like an app error. People needed to know whether money had moved and what to do next.

Banking jargon slowed people down

People thought in terms of 'send', 'scan' and 'request', not account terminology.

04 · Decision stories

The calls that shaped the product.

Each decision follows the same path: signal, insight, options, decision, trade-off, outcome.

Decision 01

Intent-led hierarchy over banking jargon

  1. Signal

    First-time users hesitated on the home surface.

  2. Insight

    People arrive with one of four intentions.

  3. Options

    Account-centric home; feature grid; or intent-first actions.

  4. Decision

    Organise the main surface around Send, Scan & Pay, Request and Manage accounts.

  5. Trade-off

    Less space for cross-sell placements on the main surface.

  6. Outcome

    Around 40% fewer initial navigation steps (reported).

Decision 02

Failures as product moments

  1. Signal

    Payment failures were read as generic app errors and drove duplicate attempts.

  2. Insight

    The key question was always: did my money move?

  3. Options

    Generic error with retry; or state-specific messaging with recovery actions.

  4. Decision

    Model the transaction states with risk and engineering, then show funds status, timing and a safe next step for each one.

  5. Trade-off

    Many more states to design and test than a single error screen.

  6. Outcome

    Clearer understanding of failures and fewer duplicate payments (reported).

Decision 03

Always-visible multi-bank state

  1. Signal

    People with several linked accounts paid from the wrong one.

  2. Insight

    The active account had to be visible at the moment of payment.

  3. Options

    Hide in settings; confirm on every payment; or a persistent account indicator with one-tap switch.

  4. Decision

    Persistent active-account indicator with one-tap switching inside checkout.

  5. Trade-off

    Extra visual weight on an already dense payment screen.

  6. Outcome

    Higher reported confidence during repeat payments.

05 · System view

Payment tokens that scaled across the super app

Shared patterns let lending, gold, insurance and bill payments reuse the same trusted payment moments.

Biometric and PIN prompts

One consistent authentication pattern across every brand.

Bank badges and account chips

Recognisable bank identity at every step of a payment.

Numeric keypad

A shared, accessible keypad for amounts and UPI PIN.

Transaction states

Success, pending, failed with funds deducted, failed without funds deducted, and reversed.

The shipped experience

Intent-first home, multi-bank management and clear payment status, on iOS, Android and web.

  • Intent-first payment home
  • Persistent bank account indicator
  • Status screens that say whether money moved

06 · Impact

What changed, and how we know.

Every figure is labelled with what it measures. Reported figures are being matched to their original baselines and periods.

~42%Fewer payment steps
Reported outcomeSteps to complete a standard payment, before and after redesign. Baseline and measurement period being confirmed.
+22%Task completion
Reported outcomeSuccessful completion of core payment flows. Baseline and measurement period being confirmed.
+33%Self-serve recovery
Reported outcomeFailed payments resolved by users without contacting support. Baseline and measurement period being confirmed.

07 · Reflection

What I'd do differently at the next scale.

I would publish the transaction-state model as a shared artefact earlier; it became the most reused thing we made.

Next time I would pair every reported improvement with a controlled comparison, so the numbers are easier to defend.