Prommt

Year
2025
Duration
2 Months
Role
Product Designer
Team
Michael Flynn - Product Lead

When a customer buys or returns a car, they shouldn't have to wait 3-5 days to see their money. Dealership staff shouldn't have to chase cheques or manually reconcile transfers that never arrived.

I led the end-to-end design of Prommt's first Pay by Bank payout product, turning NatWest's open banking API into a tool dealership staff could use in under a minute and customers could trust on first contact.

• The Result: Customers left the dealership with funds already in their account. 3-5 day delays became instant transfers.

The Scale: Prommt processes over €3B in transactions across its platform. Pay by Bank now accounts for 85% of payout volume.

The Impact: What started as a pilot feature became the default way dealerships send money.

Prommt

Year
2025
Duration
2 Months
Role
Product Designer
Team
Michael Flynn - Product Lead

When a customer buys or returns a car, they shouldn't have to wait 3-5 days to see their money. Dealership staff shouldn't have to chase cheques or manually reconcile transfers that never arrived.

I led the end-to-end design of Prommt's first Pay by Bank payout product, turning NatWest's open banking API into a tool dealership staff could use in under a minute and customers could trust on first contact.

• The Result: Customers left the dealership with funds already in their account. 3-5 day delays became instant transfers.

The Scale: Prommt processes over €3B in transactions across its platform. Pay by Bank now accounts for 85% of payout volume.

The Impact: What started as a pilot feature became the default way dealerships send money.

No More Processing Fees

Account-to-account transfers removed card processing fees entirely.
Merchants kept more revenue from every transaction without changing how they work.

Funds in Seconds, Not Days

Customers left dealerships with funds already in their accounts.

What previously took days now completed while they were
still on-site.

85% Adoption Rate

What started as a pilot became the default way dealerships send money.


85% of transactions are now utilizing the Pay By Bank feature.

Refined Design System

Through design system refinement, we accelerated ideation and development by 40%

What used to take weeks now takes days.

The payout feature, I worked on was presented
@Automotive Management Expo in Birmingham, UK.

The payout feature, I worked on was presented @Automotive Management Expo in Birmingham, UK.

The payout feature, I worked on was presented
@Automotive Management Expo in Birmingham, UK.

Opportunity

Combat Rising Fraud

1 in 5 UK dealerships were hit by payment fraud in 2024, with chargeback losses running into thousands per incident. Card payments had no reliable authentication layer. Open banking flipped that. Customers verify through their own bank, which means the identity check happens at the source, not a third-party form.

Merchants Could Accept Money. They Couldn't Send It.

Dealerships had inbound payments covered. Refunds and trade-in payouts were a different story: cheques, manual bank transfers, and 3-5 day waits. There was no digital outbound flow. Adding Pay by Bank closed that gap and made Prommt a complete payment platform, not just a collection tool

Regulations Were Here

EU and UK regulators were already pushing instant account-to-account payments as the new standard. Building Pay by Bank ahead of that curve meant Prommt wasn't reacting to compliance requirements, it was already there. That positioning mattered to the banks and enterprise clients evaluating the platform.

EU and UK payment regulations are heavily promoting instant account-to-account transfers.

Prompt positioned ahead of compliance, building credibility with banks and attracting enterprise clients

Payment Solutions

Provide merchants both inbound and outbound payments

Merchants could accept payments but not send them digitally. Offering both created a comprehensive platform increasing merchant retention.

Challenges

API Complexity

NatWest's API required three distinct data structures with precise banking credentials. One malformed field meant a silent payment failure with no clear feedback for the user. The design challenge was making that feel simple without hiding the stakes.

20+ Statuses, Zero Context

The API returned over 20 transaction statuses written for engineers, not dealership staff. Terms like payment_settled and PIN_LOCKED needed human translations and clear next actions. I mapped every status to a user-facing label before touching a single screen.

The Trust Handoff

Completing a payout required users to leave Prommt, authenticate through PayIt, then return. At the most critical moment in the flow, users were on someone else's product. Visual continuity and messaging had to work hard enough that the transition felt invisible.

Discovery

Before designing anything, I needed to understand what NatWest's API could actually do and where its limits would constrain the experience. The merchant workflow problem was already visible. The harder question was whether the technical infrastructure could support a flow that felt simple enough for dealership staff to use without training.

I worked through the API documentation and UK open banking regulatory requirements to map what was possible, what was mandatory, and what would create friction if handled wrong. Three things became clear early: the authentication flow had non-negotiable steps that couldn't be shortened, the status response system was built for engineers not end users, and the handoff to PayIt was unavoidable.

That last point shaped everything. The solution wasn't going to be seamless in the traditional sense. It was going to ask users to leave the product at the most sensitive moment. The design challenge wasn't simplifying the flow. It was building enough trust that users would follow it anyway.

Research

Method

I mapped out everything needed for integration:

  • Authentication flows

  • Required data fields

  • How statuses progress through the system

  • Business rules

  • Reconciliation protocols.

The API documentation gave me the technical picture. What it didn't give me was a user experience. NatWest's system returned 20+ statuses, required three distinct data structures, and used backend field names that meant nothing to dealership staff.

Before opening Figma, I needed to define the translation layer:

What every status meant in plain language

Translated 20+ complex API codes into human-readable labels; for example, mapping payment_settled to "Payout Success" so merchants could track transactions at a glance.

Data Abstraction

Replaced backend technical fields (e.g., merchant_id, brand_id) with intuitive business inputs like a "Location" dropdown, ensuring that non-technical would understand.

Edge Case Handling

Identified customer-facing technicalities like PIN_LOCKED that required no merchant UI but were mapped to a Transaction Detail View to empower staff with a transaction timeline & troubleshooting logic.

A dealership staff member saw a standard form. A customer saw three screens on their phone. Neither knew they were moving through an OAuth flow, webhook confirmations, and real-time status polling.

That gap between what the API required and what users experienced was the design work.

Workflows

Merchant Side: Creating Payout

Everything in the Research section had one job: make this invisible. The merchant flow had to feel like sending a message. The customer flow had to feel like receiving one.

Underneath both was an OAuth handoff, real-time status polling, and a three-party authentication system. None of that needed to be the user's problem.

Dealership staff access the payout form and input:

  • Amount

  • Secret answer

  • Customer details

  • Delivery channel (email/SMS/link).

  • They can customize the message or use the default provided.

  • Review and send.

Secret Answer Requirement

Merchants create a security question when initiating the transaction, and customers must answer correctly before accessing funds.

This prevents unauthorized payout claims.

Form Validation

Form validation happens when merchants submit rather than during data entry.

This reduces interruption and allows merchants to complete the entire form at their own pace before receiving feedback on any errors.

Emphasis on Simplicity

Merchants input amount, customer details, delivery method & security question.

The streamlined form mirrors Prommt's payment workflows, reducing learning curve for existing users.

Customer Side: Receiving the Payout

Customers receive branded notifications via
SMS, Email, or link) with transaction details, security answer and a link to complete the payout.

  • Customers enter their security answer

  • Select their bank from 300+ options via PayIt

  • Authenticate through their bank's portal

  • Choose Account & receive funds instantly.

Transparency = Trust

Notifications explain exactly what customers need to do and why. Each screen reinforces security through clear messaging & branding about the merchant, transaction amount, and verification purpose.

Customers always know what's happening and what comes next, reducing anxiety around a new payment method.

Authentication & Security

Customers authenticate through their own bank's portal, not a third-party login screen. PayIt connects to 300+ UK banks, allowing customers to use existing credentials in a familiar interface.

No sensitive banking details are shared with merchants, maintaining security and privacy.

Instant Gratification

After verification and bank selection, funds arrive in customer accounts instantly. Customers leave dealerships with money already transferred:
- No check deposits
- No 3-5 day waiting periods
- No trips to the bank

Real-World Prototype

Real-World Prototype

Utilizing my experience with design LLMs such as: Cursor, V0, Lovable & Figma Make, I built a prototype to showcase the Payout feature from a merchant's perspective.

This prototype was shown to investors & prospective new clients at the AM Live Expo and showcased the Payout feature being used in a real-world setting

This prototype was presented at AM Live Expo to investors & prospective new clients and showcased the Payout feature being used in a real-world setting

Have a play with it yourself. It's only functional for desktop

Have a play with it yourself.
It's only functional for desktop

Have a play with it yourself. It's only functional for desktop

*Opens Figma

*Opens Figma

Design System Improvements

The Problem

Prommt's design system worked until you tried to use it somewhere new. Data tables had to be manually rebuilt for every context. Tab selectors duplicated every time a new view was needed. Components were detached, renamed, and left to drift. Nomenclature was inconsistent enough that finding anything required asking someone.

I audited every component, fixed the architecture, and shipped it alongside the payout feature.

The Solution

Robust Components

Button groups, tab selectors, and data tables built with variants and booleans. Swap states, adjust layouts, change content without detaching from the master component.

Rapid Ideation

The payout table fully designed across every state, empty, loading, confirmed transaction, error, and sidepanel without rebuilding or detaching a single component.

Organized Design System

Applied consistent naming and syntax across every component so designers can find what they need via Figma's search, from any file, without opening the design system directly.

The Impact

Feature iteration and delivery accelerated by 40%.

This efficiency gain compounds across every subsequent design excercise, feature & established reusable patterns for future products at Prommt.

Key Takeaways

→ Technical constraints are a design problem

NatWest's three-part data structure and 20+ API statuses weren't engineering concerns to hand off — they were the design brief. Translating them into a merchant experience that felt simple was the core challenge.

→ Security becomes trust when you frame it right

The secret answer requirement could have felt like friction. Positioned correctly, it became the feature that made merchants confident enough to adopt.

→ Information Architecture

Organizing 20+ statuses by "what users need to do" rather than backend processes made the system understandable to non-technical dealership staff.

→ Design Systems are product infrastructure

Fixing component architecture mid-project is expensive.
The 40% efficiency gain wasn't a bonus it was the cost of not having it right from the start.