Before, a list with no hierarchy. After, what matters comes first. It's a mockup in a device, and the new profile shipped in 2024.
In 30 seconds
- The app's profile held orders, credit, bonus points, loans and settings with equal weight. The financial products that set Bemol apart were as buried as the items almost nobody opened, and every team had a different hierarchy in mind.
- I ran interviews and card sorting with six areas to reach an order everyone helped produce, and redesigned the profile's architecture and visual hierarchy, usability-tested before it went to production.
- Three months after launch, against the same period before: +170% use of Loans, +153% of Crediário and +126% of My Account and Bonus History.
- Bemol Digital
- Senior Product Designer
- Product Design · Research
- Bemol App
- Launched · 2024
What changed
Three months after launch, measured against the same period before.
- +170%
- Loans section usage
- +153%
- Store credit usage
- +126%
- My Account and Bonus History
The profile had become a drawer
Profile had accumulated features for years without anyone deciding what belonged there. Orders, credit, bonus points, loans and settings sat side by side with equal weight, in an order nobody had chosen.
The result was that high-value features (the ones the business most wanted used) were as buried as the ones almost nobody opened.
Three problems of hierarchy
Every item had the same visual weight
Purchase tracking and account settings were presented identically. With nothing to signal importance, the list read as a wall and people scanned rather than navigated.
Financial features were invisible where they mattered
Crediário, bonus points and loans are the products that make Bemol different from a generic retailer. In the profile they were indistinguishable from a settings entry.
Nobody agreed on what mattered most
Different teams had different mental models of the hierarchy. Without a shared ordering, any redesign would just be one opinion replacing another.
Settling the hierarchy with evidence
Card sorting was the piece that ended the argument.
Four artefacts, in order. Each one answers a question the previous one left open.
Management, QA, engineering, logistics, trade marketing and support. They surfaced how differently each area read the same screen.
Turned opinion into a ranking that could be defended. It replaced "I think orders should come first" with a result everyone had participated in producing.
Showed which conventions customers already carried in before they got here.
Tested the ranking before it went to production.
Test evidence, annotated: the findings that adjusted the final hierarchy.
Card sorting didn't tell me the right order. It told six people who disagreed that they had produced one together.
The redesigned profile
The new order follows the card sorting and aggregate usage, and every item stays in the same place. Crediário got its own emphasis.
The product that most differentiates Bemol, now with its own weight on the screen.
What didn't ship
Ordering the profile by individual usage
Personalising per customer would have made the screen unlearnable: the item you used yesterday moves tomorrow. We ordered by aggregate behaviour and kept the position stable.
Merging bonus points and store credit
They look similar and behave differently. Combining them tested badly: people expected one balance and found two rule sets.
Who did what
- Interview script and facilitation of all six sessions
- Card sorting design and analysis of the resulting hierarchy
- Information architecture and visual hierarchy of the new profile
- Usability test script and post-launch measurement plan
- Usage data extraction, with the analytics team
- Implementation, with the core app team
- Prioritisation of which features to highlight, with product
Reflections
The redesign changed almost nothing visually.
No new component shipped with this. What moved the numbers was ordering and weight: deciding what comes first and what gets emphasis.
That's also why the result travels: the +170% on loans came from the feature finally being where people were already looking, and that holds for any other.