Centiiv Escrow

Led by Folasewa

Designing Trust into Cross-border Payment Escrow

Centiiv bridges fiat and stablecoin payments for global merchants and liquidity providers. I designed the escrow checkout flow —
the product's core trust mechanism — from zero to launch.

Role

Product Designer | UX Designer
Owned: IA strategy, UI Design, Design system, Engineering handoff

Company

Centiiv

Industry

Fintech | Payment

Outcomes
20M+

transaction volume

10% faster

dev integration via design system

3 rail

fiat, card, and stablecoin unified under one checkout surface

0 → 1

product built from scratch — no existing design system or patterns

THE PROBLEM

Global payments break without a neutral holding layer

In cross-border B2B payments, money moves between merchants, aggregators, and customers across different currencies and time zones. Without escrow, either party holds the risk during transfer — merchants fear non-delivery, customers fear non-receipt.


Existing solutions forced users to trust the platform blindly or rely on email follow-ups. Centiiv's escrow USP — hold funds until both sides confirm — needed a UI that made that trust tangible and frictionless across fiat, card, and stablecoin rails.

01
Research
I led the field research and accessibility testing to identify painpoints
Research
02
Iterations & Design
03
Testing
04
Design handoff

THE SOLUTION

Escrow flow (4 critical states)

Customer info + terms

Email capture + escrow T&C before payment. Sets legal context early.

Pay — 3 rails

Card, bank transfer (expiring virtual account), or stablecoin. One surface.

3.

Dispute filed

Funds lock. Merchant submits evidence. Structured, not chat-based.

4.

Propose settlement

Either party proposes terms. Other party accepts, rejects, or counters

KEY DESIGN SCREENS

Design Decisions Made

Personal info + escrow entry

  1. Decision : collect only email at step 1. Reduces drop-off before payment. Escrow T&C surfaced here, not buried in footer.


  1. Decision : all 3 rails on one screen — reduces abandonment from method confusion.

Card payment + order summary

Decision: escrow fee (0.6%) shown transparently in order summary. Prevents post-payment confusion and builds trust in the fee model.

Stablecoin rail

Decision: show wallet address + exact USDC amount with expiry — crypto users need precision

Bank transfer (escrow account)

Decision: expiring virtual account per transaction — eliminates misrouted payments

Merchant verification

Decision: binary CTA — verify OR dispute. No ambiguity. Tied to 24hr SLA.

Dispute transaction page

Decision: structured form (reason category + description + evidence upload) instead of open chat. Forces specificity, speeds resolution. Escrow lock notice shown prominently.

Create settlement proposal

Decision: separate proposal flow from chat. Merchant and customer can negotiate amounts without losing dispute history. Counter-proposal pattern modeled on legal ADR.

WHAT DIDN'T WORK

Key Trade-offs Made
I dropped

Persistent merchant bank account. Simpler to build but caused payment attribution failures in testing — unacceptable at this transaction volume.

I chose

Single expiring virtual bank account per transaction. Prevents mis-routing; adds backend complexity but removes a major support category.

I dropped

Inline dispute form on the same screen. Tested it — merchants skimmed past it. Separating confirm modal improved completion rate in usability sessions.

I chose

Structured dispute form with evidence upload. Non-negotiable for legal validity. Users initially wanted chat-only — usability tests showed chat led to ambiguous outcomes.

AI INTEGRATION

How I used AI in this project
Edge case mapping

Used Claude to stress-test the dispute state machine — caught a missing state (expired escrow, no merchant action) before dev handoff.

Microcopy

Generated 5 variants of the escrow expiry warning. Chose the one that tested clearest with non-crypto users in a quick 5-person session.

REFELECTIONS & LEARNINGS

What I'd do differently

I would push earlier for quantitative data on dispute resolution time. I designed the 24hr SLA based on competitor research, but without baseline data from Centiiv's own merchant behavior, we were guessing. I would also add a progress indicator between payment submission and escrow confirmation; users in testing were unsure whether funds were "held" or "sent."

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