Universo (Sonae) · 2021
In-Store Finance Solution (BNPL)
An end-to-end BNPL product that opened in-store credit at Sonae’s stores to everyone, not just Universo Card holders. Within three months of launch, conversion rose 45%, beating the 30% target.
- Role
- Product Designer
- Company
- Universo (Sonae Group)
- Timeline
- Nov 2020 – May 2021
(~ 7 months) - Core Team
- PMO, Product Manager, Project Manager, QA Lead, IT Architect
- Surfaces
- Store onboarding, Store backoffice, Client area, Ops backoffice
- Tools
- Adobe XD, Miro, Azure DevOps, Notion
Context
Universo, Sonae Group’s financial services brand, lets customers buy at thousands of Sonae stores (retail, electronics and more) and pay later in instalments. In late 2020, Universo wanted to evolve this in-store credit into a standalone product.
The problem
The existing in-store finance solution was only available to Universo Card clients. Everyone else was left out at the till, which capped adoption and the business behind it.
Main objective
Research, strategise and design an end-to-end BNPL credit solution for Sonae’s stores, giving people in Portugal access to low-barrier credit, and increase conversion by 30% over the card-only solution.
My role
I was the only designer in a squad with a PMO, Product Manager, Project Manager, QA Lead and IT Architect. I owned research, UX and UI across four connected surfaces, from the first stakeholder interviews to developer handoff, in a company that was new to user-centered design.
Approach
With tight deadlines and pressure to start building, I used a Lean UX approach: iterative, collaborative and evidence-based, with a pause to reflect after each cycle. The aim was to get feedback as early as possible so the squad could make quick, confident decisions.
- 01
Understand
Open-ended interviews with subject-matter experts, whiteboard brainstorming and frequent squad sessions to understand the product, the business expectations and each type of user.
- 02
Explore
User flows, edge-case discovery, sketching and wireframing, design critiques and lo-fi prototypes, iterated with the PMO and PM.
- 03
Materialize
Hi-fi prototypes, guerrilla usability testing and close handoff with engineering.
What we learned
Three audiences, and the business behind them, would use the product with different needs:
Store staff
A fast, intuitive onboarding that guides them through each step, saves progress automatically, and gives them aftersales actions like disbursing and refunding.
Operations teams
Tools that feel close to what they already use, with full aftersales capabilities: disburse, refund, KYC, irregularities and logs.
End clients
Easy access to credit for bigger purchases, then a simple self-service area to follow their payment plan, download legal documents and keep their details up to date.
The business
An independent product integrated with Sonae’s systems, credit simulation, contracts showing the payment schedule, and the same backoffice access for existing users, built in-house on Angular.
Constraints
- Tight deadlines, limited time for user research, and a waterfall-leaning culture.
- Reusing APIs from a previous solution that wasn’t built with user-centric principles.
- Stakeholders unfamiliar with UX and sceptical about the value of design thinking workshops.
Workflow diagrams
Before designing any screen, I mapped every flow end to end with the squad, covering the happy path and the edge cases. These diagrams became the shared reference for business and IT across all four surfaces.
- Onboarding flowFrom the store backoffice through KYC, credit simulation and the decision (approved, refused or counter-proposal) to signature and disbursement.
- Homebanking sitemapClient login with OTP, password recovery, and the client area: purchases, personal details, consents and profile.
- Refund credit flowHow operations agents find a credit application and refund a partial or total amount.
- Disburse credit flowHow operations agents disburse credit, filtered by application or VAT number, with client OTP confirmation.
- Impersonation sitemapThe client area as operations agents see it, after the client confirms with an OTP.
The solution
Four connected surfaces covering the whole credit lifecycle, from the moment a customer asks for credit at the till to their last instalment.
01 - Onboarding platform for stores
- Guided, stepped onboarding redesigned from the old card flow, with autosave as a fallback for busy tills.
- Credit simulator whose inputs and logic were iterated with the PMO and PM.
- Simpler KYC, reworked from the Universo Card’s old solution.
- Three clear outcomes: approved, refused or counter-proposal, with error prevention before a proposal is rejected.
02 - Backoffice operations for stores and ops
- Product segmentation, so the new credit lives alongside existing products.
- New flows to disburse credit, disburse pending credit and issue refunds.
One business goal was to run all Universo credit operations from one place. The Universo Card was already managed in this backoffice, so In-Store Credit joined it there.
New tabs let agents switch between products, from Universo Card to In-Store Credit.
New flows were added to perform new operations:
- Disburse pending credit: A queue of approved credit waiting to be paid out. Agents find each application by number or VAT number, and the customer confirms with an OTP.
- Refunds: Agents find a credit application and refund part or all of it, for example when a customer returns a purchase.
03 - Client area for end clients
- A central payment plan component: plan details, update IBAN, download contract and pay early.
- Generate a Multibanco reference in a few steps and receive it by SMS.
- Account activation with OTP, payment confirmation emails, and personal data updates.
- A first-login guided tour to help new clients find their way.
First-login guided tour
Early repayment
Update IBAN
04 - Client area backoffice for ops teams
Agents see an impersonated view of the client area, exactly what the customer sees. That makes customer service calls faster: agents can act for the customer and send payment plans, contracts and Multibanco references by email or letter.
Key design decisions
- 01
A “no” isn’t the end
Treating the counter-proposal as a first-class path, alongside approved and refused, gave customers who didn’t qualify for their first request a way forward, instead of losing them at the till.
- 02
One component, two audiences
The client area and the ops backoffice share the same components, so agents support clients from the exact view they see, and engineering builds each piece once.
- 03
Familiar over novel for staff
Store and ops tools stayed close to what teams already knew, so teams could adopt the new product with little training.
- 04
Design for interruptions
Autosave and a clear stepper protect progress in a noisy store, where a sale can be interrupted at any moment.
Validation & handoff
With little time for formal research, I ran guerrilla hallway usability tests to validate design decisions, catch usability issues fast and uncover opportunities, a good fit for small, frequent iterations.
For handoff I shared CSS specs in Adobe XD, assets and interaction descriptions in Azure DevOps, and joined weekly meetings with engineering.
Outcomes
The counter-proposal path alone kept customers who’d otherwise have left without credit.
Beyond the numbers, it also helped shift how the company saw design, from a final polish to part of building the product.
Reflections
Challenges
- Convincing stakeholders of the value of product design, as the only designer in a waterfall-leaning company.
- No room for more extensive research with real end users.
- No data scientists on the project, so little data, KPIs or measurement.
- A fast time-to-market that left little time for testing and iteration.
What I’d do differently
- Test more, with a more methodical usability testing approach.
- Define how to measure each design decision and its impact from the start.
- Create a design system to scale across the four surfaces.
Got questions?
I’m always excited to collaborate on innovative and exciting projects.