Role
Lead Product Designer
with external design partner
with external design partner
CASE STUDY · DIGITAL IDENTITY
From a hidden icon to an upfront login
I redesigned Digidentity's login and account switching, replacing an undiscoverable switch icon with a carousel that shows users every account at a glance. With no user testing available on a regulated product, I drove the decision with workshops and a structured survey, and aligned the final design with Apple's Liquid Glass as it launched mid-project.
Context: Digidentity is an eIDAS Qualified Trust Service Provider on the EU Trusted List, whose platform has created 25M+ verified identities across 180+ nationalities.
The Discovery
The Problem
Multi-account users couldn't find the account switcher. It lived behind a small, unlabelled icon in the top-right corner, invisible to new users and frustrating for returning ones. Login simply defaulted to whichever account was last active, with no way to see the others unless you knew to hunt for that icon. In a digital identity app, logging into the right account — personal vs. business, different credentials — has real consequences.
On top of that, the login itself felt dated. It hadn't been refreshed in years and didn't reflect the trust and professionalism expected of a product that handles government-issued identity.
The opportunity: How might we…
- surface every account right at login, without overwhelming people who only have one?
- make account switching obvious to a first-time user, not a hidden gesture?
- modernise a dated government-identity login so it feels trustworthy and current?
Who I Designed For
Existing users with multiple accounts and credentials, typically EU-based professionals who hold both personal and business identity products (eHerkenning, eSGN) and switch between them regularly. The design also had to stay clean and simple for single-account users, who shouldn't pay a complexity tax for a feature they don't need.
Constraints I Designed Within
01No direct user testing available: recruiting external users to test a regulated identity product isn't straightforward, so I had to validate another way.
02The design had to gracefully handle 8+ accounts, blocked and inactive states, dark mode, and tablet sizes.
03It had to feel native on both iOS and Android while staying consistent with Digidentity's existing component library.
Strategy & Logic
Three Competing Directions
Rather than commit to one idea, I designed and prototyped three: a Small Carousel, a Big Carousel, and a Dropdown. Each made a different bet about how visible account switching should be, and each was mapped against five core login scenarios.
Workshops Instead of a Vacuum
Because I couldn't run external user tests, I ran cross-functional workshops with the people closest to the product — the CTO, service desk, product managers, and developers — to pressure-test each direction against real user pain, technical reality, and support burden.
The Pivot
Stakeholders were split on which direction to take. With no user testing on the table, the decision was at risk of being settled by whoever argued hardest: exactly the wrong way to choose in a trust-critical product.
With no user testing on the table, I replaced opinion with evidence: a structured survey, not the loudest voice in the room.
So I ran a structured internal survey using the prototypes, asking participants to complete real tasks and rate each option screen by screen. The result was clear: the Carousel beat the Dropdown overall (7 of 12, 58%) and won decisively on the screens that mattered most: seeing every account at once, and discovering "Add account" at 6 of 12 (50%). The Dropdown edged ahead only on single-account login, where there's nothing to switch between. That gave the team a defensible, data-backed direction instead of a stylistic argument.
There was a second pivot outside my control: the project paused for reprioritisation and restarted in mid-2025, right as Apple announced Liquid Glass at WWDC. The translucent direction I'd already been exploring suddenly became the platform standard, so I leaned into it rather than fighting the tide.
Aligning with Liquid Glass
I brought Liquid Glass-inspired treatments into the final design (translucent headers and subtle refraction on the account cards) so it felt native to iOS 26 while staying true to the Digidentity brand. For Android, I adapted the visual treatment to feel native to Material Design while keeping the same interaction patterns and hierarchy.
The Execution
The Final Login
A Big Carousel of swipeable account tiles. Open the app and you see your accounts as prominent cards — each showing the account name, company, and email — with the current one centred. Swipe to switch, enter your PIN, done. The flow is immediate and visual, with zero hunting for a hidden icon.
Every State, Designed
01Single account: a clean, focused login with no unnecessary carousel UI.
02Multiple accounts (up to 7): the horizontal swipe carousel.
038+ accounts: a graceful fallback to a scrollable "All accounts" list.
04Blocked, inactive, and add-account states: each with a clear next step, plus a full dark-mode treatment.
Design System & Handoff
The account tile shipped as a reusable component with variants for active, blocked, inactive, and selected states, and I updated the PIN pad and login flow to match. Everything was handed off with detailed iOS and Android specs: interaction details, edge-case documentation, responsive behaviour, and prototype flows developers could step through for every scenario.
The Conclusion
Results
Since launch, wrong-account login attempts are down 60% and successful multi-account switches are up 40%, both confirmed by the analytics team. Worth noting: many users only open the app once a year around tax season, so logging into the wrong account wasn't just an annoyance, it was a real blocking problem. The redesign puts every account in front of the user from the first second, on a product where logging into the right identity actually matters.
Future Development
This is Phase 1 of a longer arc. Phase 2 introduces customisable account icons, so multi-account users can recognise accounts at a glance, moving the product from "functional" to genuinely personal.
Phase 3, in design now, extends biometric unlock into the wallet itself, opening credentials with Face ID, building on the Face ID login explored in earlier prototypes. The goal: faster access without another PIN step, with the assurance model intact.
Reflections
The workshops and survey were a strong substitute for user testing, but not a full replacement. With more time, or in a less regulated setting, I'd pair the survey with real usability sessions to validate the carousel with actual multi-account users before handoff.