How removing choices from a screen — not adding them — turned around a declining conversion rate.
COMPANY
MFSG
YEAR
2025
Overview
The Loan Protection Plan (LPP) is an optional add-on Money Mart customers can purchase alongside their loan. For a small additional cost, it protects them if they lose their job, get into an accident, or face another major life event that makes repayment difficult. It's a product that genuinely helps customers in vulnerable situations, and it's also an important revenue line for the business.
Following a major app redesign, LPP attach rates collapsed. Between Q4 2025 and Q1 2026, monthly opt-in rates hovered between 7% and 15% — well below historical performance, bottoming out at 7% in December.
Leadership's initial response was to revert entirely back to the previous app design. I was brought in as the sole designer to figure out what actually went wrong and fix it — fast.
What We Overcame & Achieved
The redesign hadn't changed the product. It had changed how many decisions customers had to make at once.
The new app consolidated loan amount, loan term, and the LPP decision onto a single screen, with interactive sliders customers could adjust in real time. On paper, that looked like more flexibility. In practice, it meant customers were trying to minimize their monthly payment and decide on optional insurance at the exact same moment — and when someone's focused on cutting costs, an added cost is the easiest thing to say no to.
The previous app version, before the redesign, had handled this differently: it showed customers their maximum approved loan amount and term by default, and made the LPP decision the one clear choice on the screen. That version consistently performed better — which is exactly why leadership wanted to revert to it.
Rather than a blind rollback, I diagnosed why the old version worked — and rebuilt that logic inside the current design system.
A full visual rollback would have meant abandoning the new app's design language entirely — undoing months of design system work to fix one screen. Instead, I compared the old and new experiences side by side to isolate exactly what made the older pattern effective:
Metric
Old app (worked)
New app (broken)
Decisions per screen
One at a time
Three at once (amount, term, insurance)
Loan terms
Defaulted to max approved amount and longest term
Customer had to adjust sliders before seeing anything settled
LPP decision
Clear yes/no choice, front and center
Competed for attention with loan calculations
The fix: keep the current app's visual design, but restructure the flow. The main screen would default to the customer's maximum approved terms and present LPP as the single, clear decision. Anyone who genuinely wanted to adjust their loan amount or term could still do so — just by tapping through to a separate screen, rather than having those controls compete for attention by default. Working with our UX writer, we also refined the LPP prompt language itself to match how our customer base actually thinks about cost and risk, rather than defaulting to generic insurance copy.
Before & After
Monthly LPP attach rate rose from 17% in May to 24% in July — a 41% relative lift.
Weekly opt-in peaked at 29% in mid-July, exceeding the original app's historical performance.
LPP policy volume nearly doubled — from 109 accepted policies in May to 214 in July — while application cancellation rates stayed low and stable (5-8%).
The Why and How
Cognitive load at the moment of commitment is a conversion killer.
The loan summary screen is the last major decision point before a customer finalizes their application — already a high-stakes moment. Asking someone to calibrate their loan amount, choose a term, and evaluate optional insurance all in that same instant is too much. Most people default to the path of least resistance: skip the extra thing, keep moving.
Defaulting to maximum terms wasn't about pushing customers toward a choice. It was about removing a decision that didn't need to happen yet.
Presenting the maximum approved amount and longest term by default meant most customers — who were satisfied with what they'd been approved for — never had to touch a slider before reaching the question that actually mattered. Anyone who wanted something different was still one tap away from changing it.
Speed was part of the brief — and part of the lesson.
This went from brief to release in under four weeks. That was possible because the fix wasn't invented from scratch — it was a return to a pattern that had already proven itself, adapted to the current design system rather than discarded alongside it. The broader lesson: sometimes the highest-value design decision isn't a new idea, it's recognizing that a previous approach worked for a specific reason, making that case with evidence, and executing the fix with precision instead of using the moment to add unnecessary complexity.
Designing high-trust products that balance compliance, clarity and conversion.
Schedule a call with Abhijeet
