Designing for borrowers who deserve better.

Reevo Money, UK. Senior product designer on the loan management app for near-prime borrowers, people just outside mainstream credit. After a late vendor change, I had a new login flow in live testing within 14 days. Users noticed: satisfaction (NPS) and retention up, missed payments and collections costs down.

near-prime lending100k+ active usersend-to-end2024 to 2026

Problem.

Near-prime borrowers were failing on experience, not finances: a desktop-only portal, no alerts, no view of their options.

Desktop only. Silent until it was serious. Stressful by design.
existing portalpre-app2024
fragmented

Locked out on mobile

Desktop-only portals and no central data hub, so borrowers got inconsistent information at the moments that mattered.

invisible

Arrears stayed hidden

Zero proactive notifications. Repayment options were out of sight exactly when a borrower hit temporary difficulty.

anxious

Dread before every due date

Support ran on reactive calls, and collections had no touchpoint before problems escalated.

the result

A system that made a hard situation worse

People who don’t qualify for mainstream credit, already under strain, left to manage a loan through a tool that worked against them.

Research.

25 interviews across the UK. The signal that cut through was tone, not features.

25

in-depth interviews with near-prime borrowers, run before a single screen was drawn.

one to oneUK wide2024
what we heard

“Most borrowers had never accessed their loan portal, not because they didn’t want to, but because it wasn’t mobile accessible. Anxiety around payment deadlines was near universal.”

pattern 01

Deadline anxiety

Every borrower wanted to know “am I okay?” before anything else.

pattern 02

Proactive, not reactive

Communication before a crisis, not crisis management after one.

pattern 03

Mobile first

Strong mobile behaviour in a group served only by desktop.

the finding

Dignity was missing

Borrowers wanted to be treated as capable adults in a hard situation, not as defaulters to be managed.

Features.

Five features, one principle: reduce the stress a borrower feels the moment the app opens.

5

core features, each one answering something borrowers said in research.

calm by defaultoptions up frontnothing hidden
01 · loan dashboard

At-a-glance loan status

Paid, left to pay, the next payment and whether you’re on track. It answers “am I okay?” in under three seconds, without navigating anywhere.

02 · payment schedule

A timeline, not a table

Upcoming payments with context on each date, in a legible interactive flow.

03 · arrears banners

Options, not panic

Compassionate, action-first alerts that surface choices. Tone tested in research.

04 · flexible payments

Repayment choices, early

Repayment choices visible in-app early, cutting support calls from early-stage arrears.

05 · smart notifications

Timely, personal reminders

Proactive payment reminders and account updates, replacing silence with relevant contact.

System.

A design system built to WCAG 2.1 AA from day one, because accessibility in financial products is a trust signal.

Beautiful mockups don’t matter if they never reach users.

The principle behind every decision on this project.

Figma librarybrand layerhandoff ready
components

Buttons, alerts, forms, cards

A full component library with Reevo branding layered over system parts, so engineering built from one source.

accessibility

Contrast, focus, errors

AA across every interactive state, not as a tick box.

tokens

Colour, spacing, type

Tokens for engineering handoff, no redlines needed.

coverage

All five features, edge states, empty states

Nothing shipped as a happy path only. Every state a borrower could land on was designed and reviewed.

Pivot.

Kobil, our authentication vendor, bailed out with the app designed and ready. We chose AWS Cognito the same day.

14d

from that decision to a redesigned login flow in live testing on AWS Cognito.

Kobil outCognito inno process
the decision

Same-day call: AWS Cognito

Authentication gates every screen, so nothing could go live without it. We stopped protecting the pixel-perfect vision, started unblocking engineering, and I took on the login flow redesign that day.

every morning

Build review on real devices

Working prototypes in hand, no simulations. Issues triaged against a critical-versus-deferrable list.

all day

Open room

Design and engineering in continuous sync. No ticket queues, decisions in minutes. The absence of process was the process.

the login flow · 14 days

Redesigned around Cognito

Cognito’s UI customisation is severely limited, so the sign-in screens couldn’t be pixel-perfect. Instead: system components with Reevo branding, contextual error help, an SMS fallback for accessibility, and design QA twice a day inside the sprint.

Results.

Post-launch, against the same borrower cohort, four measures moved the right way.

Users noticed. So did the collections team.
before and afterreal cohortmeasured
customer NPS
↑

Double-digit NPS gain

Active app users versus non-app borrowers in the same cohort, tracked by in-app survey.

retention
↑

Renewals up about a tenth

App users renewed more than non-app users against the pre-app baseline.

missed payments
↓

Missed payments down about a tenth

Driven by smart notifications and the visible payment schedule.

collections
↓

Collections a few percent cheaper

Faster arrears resolution from early in-app intervention, measured against collections team data.

next case study

Component Library

The MCF Portal design system, built from scratch as senior designer for 500,000+ users.

Design systemRead case study

Talk.

Open to senior and lead product design roles in India or remote.

Still reading? Good. Let’s build something that actually works.
NoidaIST
email

Write to me


Or email me at vaibhav.bn@gmail.com

linkedin

in/vaibhavbn

Work history, recommendations and posts.

Open
resume

Two pages, PDF

PDF, opens in a new tab.

View

Create a free website with Framer, the website builder loved by startups, designers and agencies.