Modena Credit Product · 2021–2025

Designing a Mobile-First Credit Product Application Flow

A deliberate architectural decision: one shared application flow that scaled from a single BNPL product in 2021 to six credit products by 2025 — without ever forking the codebase.

Interactive
Small Loan
ENG
Loan Amount
€
Loan Period
months
Monthly Payment
85.15 €
Exact monthly payment will be determined during the application.
Identification
ENG
By using the service, I agree to the Privacy Policy and Terms of Use.
Contact Details
ENG
By using the service, I agree to the Privacy Policy and Terms of Use.
I agree to receive marketing messages.
Bank Selection
ENG

Select your home bank

To get the best offer, we need data from your home bank.

Swedbank
SEB
LHV
Coop
Citadele
Card payment

How does data sharing work?

We have received your account information.

If you also use another bank for important daily transactions and want to share that information with us, press 'Add another bank.'

swedbank.ee/login
S Swedbank

Log in to your bank

Modena is requesting access to your account data to evaluate your loan application.

Your credentials are sent directly to Swedbank. Modena does not store them.

swedbank.ee/consent
S Swedbank

Share account information

Modena is requesting permission to access the following data from your account:

Account balance and transactions
Income and regular payments
Existing loan obligations

Access is granted for a single assessment only. Your data will not be stored beyond the evaluation period.

Financial Info
ENG

Income

€

Obligations

€
€
Small Loan
ENG
Modena
Modena
Payout to account 3 000.00 €
Payment schedule 36 months
Monthly Payment 85.15 €
First payment
Contract and information sheet

Notifications will be sent to your email and phone.

Loan approved!

Transfer to account will be immediate.

4902

Check the code on your device and enter PIN-1

2145

Check the code on your device and enter PIN-2

Timeline
4 Years
Role
Product Designer
Team
PM / CEO
Systems Architect
Developer
Tools
Figma

Outcomes

€16.8M
Contracts Financed
6
Credit Products
+2
Screens to Launch Business Loans
<2 min
Consumer Application Time
1
Shared flow, never forked

The Problem

The real challenge wasn't building one product — it was building an architecture that could absorb whatever came next.

In 2021, Estonian merchants had no way to offer deferred payments at checkout. Customers who couldn't or wouldn't pay the full amount upfront either bought or left. Modena's BNPL changed that — 0% interest, 1–3 month terms, merchant-funded — but it was just the starting point.

The real challenge wasn't building one product. It was building an architecture that could absorb whatever came next — without rewriting the flow each time a new product launched.

BuyPlan MVP — first screen, 2021
BuyPlan MVP — last screen, 2021
The original 2021 MVP — BuyPlan, four primitive screens. Real merchant clients, just very few purchases a month.

Four Years, Six Products, One Flow

Four years of compounding design decisions on a single application flow.

Most case studies show a snapshot. This is a longitudinal story — four years of compounding design decisions on a single application flow.

I joined Modena in early 2021 and reworked the initial BNPL flow. Over the next four years, the product expanded into six credit products: BNPL, hire-purchase, small loan, consumer credit, business loan, and business credit. Each had different requirements — purchase-based products start with a merchant-set amount; loans let the customer choose. Income verification is required for loans but not BNPL. Business products add beneficiary steps and bank statements.

The deliberate decision was to never fork the flow. Instead, I designed a single branching architecture — shared screens with steps added or removed per product type. Early choices about step order, data collection points, and conditional logic meant that each new product could ship on top of what existed. The loan portfolio this generated (€16.8M) later justified building the Investor Portal to fund it externally.

Flow Overview

One flow, four branching paths, zero forks.

The interactive prototype above shows the small loan product. The full flow has 8 steps — tap through the prototype to experience it. Here is how the branching works across product types:

Step BNPL Hire-purchase Loans Business
Calculator Fixed Fixed Adjustable Adjustable
Identification ✓ ✓ ✓ ✓
Bank verification — ✓ ✓ ✓
Income & expenses — ✓ ✓ ✓
Beneficiary info — — — ✓
Review & sign ✓ ✓ ✓ Manual review

Same codebase, same design system. Steps are conditionally shown or hidden based on product type.

Key Design Decisions

Three decisions that held up for four years.

Branching architecture over separate flows.

The highest-leverage decision was resisting the pressure to build a new flow for each product. Instead, I designed a single branching system — shared screens with conditional steps. When business loans launched in 2024, it meant adding two screens, not rebuilding from scratch.

Familiarity as trust.

Smart-ID and direct bank integrations (no third-party open banking) made authentication feel routine, not risky. Estonian users already trust these systems — leveraging that reduced friction at the highest-stakes moments in the flow.

Consistent interaction patterns.

Bottom-anchored CTAs on every screen, a 6-segment progress bar for orientation, and predictable screen transitions. These patterns stayed constant as the product range grew, so returning users never had to relearn the flow.

Results

From one BNPL product to six credit products generating €16.8M in financed contracts.

From a single BNPL product with limited traction to six credit products generating €16.8M in financed contracts. Consumer flows complete in under 2 minutes; business loans take 2–3 minutes with a manual review stage. The portfolio grew large enough to justify its own external funding channel — the Investor Portal.

What I Learned

Design architecture compounds the same way code architecture does.

Four years on one flow taught me that design architecture compounds the same way code architecture does. Early decisions about step order, data collection points, and branching logic either accelerate or constrain every product that follows. Getting those foundations right meant new products shipped on top of what existed. Getting them wrong would have meant rebuilding from scratch — repeatedly.