← Swati PriyaLet’s talk
01Fintech · Education payments· 2023 — 2026

Fee payments parents finish in one sitting

A monthly fee-financing flow for parents who abandon the moment a form feels like a loan application.

Role
Senior Product Designer — owned end-to-end flow, design system contributions
Team
1 designer, 1 PM, 5 engineers, risk & ops partners
Platforms
Responsive web, Android-first mobile web
Timeline
2023 — 2026
GrayQuest fee-payment flow screens on mobile and desktop

Outcomes

What changed

  • 2M+

    parents reachable

  • −41%

    drop-off at KYC

  • 3 steps

    down from 7

Context

The situation

GrayQuest lets parents pay school fees monthly instead of in one lump sum. The product is financially regulated, so the flow carries KYC, credit checks and repayment consent — three things parents never asked for and rarely understand.

The existing journey read like a loan application: seven screens, bank vocabulary, no visible progress, and errors that appeared only after submitting. Most parents who dropped off never came back.

The problem

Parents dropped off at KYC and repayment-plan selection: too many unexplained steps, no sense of what they had already cleared, and jargon that read like a bank contract.

Research

What I learned first

12 parent interviews across 4 cities

Parents were not price-sensitive so much as trust-sensitive. The question behind every hesitation was 'what happens to my child's admission if this fails?'

Funnel + session replay audit

Two-thirds of abandonment sat in KYC document upload and plan selection. Replays showed repeated pinch-zooming on plan tables and multiple retries on the same upload field.

Support ticket clustering

Tickets grouped into three themes: 'is my payment received', 'which document counts', and 'what is the total I actually pay'.

Process

How it came together

  1. Mapped the journey against the regulation

    Wrote the flow as sentences before drawing screens, marking which steps were legally required, which were internal preference, and which could be deferred until after the first payment.

  2. Collapsed seven steps into three

    Grouped required inputs into Identity, Plan and Confirm. Anything optional moved into a post-approval profile nudge, cutting the pre-commitment ask by half.

  3. Rewrote every string in parent language

    Replaced 'tenure', 'principal outstanding' and 'processing fee' with plain phrasing and a single 'you pay in total' figure that never changed position on the screen.

  4. Prototyped and tested with real documents

    Tested the upload step with parents using their own Aadhaar and fee receipts on their own phones — which surfaced glare, crop and file-size failures a clean prototype hides.

Key decisions

Calls worth defending

Progress that counts backwards, not forwards

Instead of 'Step 2 of 7', the flow says what is left. Fewer remaining items reads as relief; a long total reads as work.

Eligibility feedback inline, before submit

Credit and document checks return as-you-go states so a rejection never lands as a surprise on the final screen.

A receipt built to be forwarded

The confirmation doubles as a shareable proof of payment for the school, because that was the parent's real end goal.

Reflection

What I’d carry forward

  • In regulated flows, the design work is mostly deciding what can be asked later.

  • Comparison tables are a desktop artefact — parents compared plans best as stacked cards with one differing line highlighted.

  • Trust came from visible status, not reassuring copy.

  • Journey mapping
  • Design system
  • KYC & onboarding
  • Mobile-first