Overview

Fivvy is a personal finance app focused on helping users optimize everyday spending through smart payment recommendations.

Based on financial goals, account status, and credit behavior, the product suggests the best payment method for each purchase.

I joined at a pre-MVP stage, where the product had no users and the value proposition was still unclear.

My role involved:

• defining core user journeys

• shaping the feature set

• designing the recommendation experience

The main success metric was conversion across the initial user journey, making clarity and speed critical.

The Problem

A powerful idea that didn’t translate into usable value

The product aimed to recommend the best payment method based on:

• rewards (cashback, miles)

• billing cycles (payment deferral)

• credit utilization (financial health)

While conceptually strong, this created a highly complex decision system that was difficult to communicate and even harder to use.

Value came too late

Users had to:

1. Configure financial goals

2. Connect accounts (and wait)

3. Input purchase details manually

4. Wait for backend processing

5. Then receive recommendations

This made the product unusable in real-world contexts, like checkout lines where decisions must happen instantly.

High friction to reach value

The experience required:

• account linking

• manual input

• multiple steps before insight

👉 The product delayed value instead of delivering it upfront.

An abstract value proposition

Unlike simpler fintech apps, Fivvy’s core idea was difficult to explain:

• dynamic payment recommendations based on financial state

This made it harder to:

• onboard users

• build trust

• communicate benefits

Technical and organizational constraints

• backend latency prevented real-time responses

• limited access to users for validation

• product decisions driven by assumptions

👉 The product risked solving the wrong problem, too late.

Understanding the System

Turning complexity into decisions

The recommendation system depended on multiple competing variables:

• rewards vs credit health vs payment timing

Each payment method could be “best” depending on context.

Structuring the system

Trough research, I mapped the product through:

• information architecture models

• domain ownership of data

• user needs vs informational value

This helped identify what actually mattered at decision time.

Reducing configuration

Originally, users had to define financial goals upfront.

I removed this requirement and replaced it with:

• automatic evaluation of all variables

• a default ranked recommendation

👉 This reduced friction and accelerated time-to-value.

Designing for speed

To avoid manual input:

• I used historical data to create predefined purchase shortcuts

• surfaced them directly on the home

👉 Users could trigger recommendations instantly.

Heuristic-based decision model

Instead of relying on slow backend computation:

• I designed a heuristic-based system

• prioritizing speed and predictability over precision

Trade-offs

• speed over accuracy

• simplicity over transparency

These decisions enabled real-world usability.

Strategy & Decisions

Simplifying to unlock value

The core strategy was:

• prioritize immediate usability over configurability

Key decision: removing goal setup

Users no longer needed to define financial goals.

The system:

• evaluated all variables

• generated a default recommendation

👉 Shift from user configuration → system guidance

Introducing financial health indicators

I added high-level indicators to:

• provide early feedback

• build trust

• reduce navigation

This required simplifying complex financial data into clear signals.

Risks

• less precision

• reduced user control

• less transparency

But necessary to deliver value in seconds.

Solution

A decision-first home

The home screen provided immediate context:

• consolidated balance

• list of payment methods

• spending overview

• quick access to recommendations

Instant recommendations

Users could:

• select a predefined purchase

• get a recommendation instantly

Ranked decision model

Recommendations were shown as a:

→ carousel of payment methods (best → worst)

Each option included:

• benefits

• drawbacks

Visual recognition system

To compare options quickly:

• bank colors + logos

• texture system for differentiation

👉 Designed for fast scanning under pressure.

Handling latency

A playful loading state:

• acknowledged system delay

• reduced perceived friction

• added personality

Key highlights

• home as a decision hub

• fast comparison via carousel

• reduced interaction cost

Impact

• reduced friction in accessing recommendations

• improved clarity of product value

• aligned experience with real usage context

The MVP launched closely aligned with the proposed solution and later iterations focused on improving clarity through onboarding and visual aids.

Learnings

Bridging product vision and real user needs

The product was driven by strong ideas, but limited validation.

This resulted in:

• unclear value perception

• delayed user benefit

• reliance on assumptions

The limits of feature-driven design

Adding features improved clarity, but didn’t solve the core issue:

• lack of deep understanding of user needs

What I would do differently

• validate the value proposition earlier

• test simpler versions before scaling complexity

• prioritize real user behavior over assumptions

Key takeaway

Designing the right product matters more than designing the product right.