2024
eXpress Settings: Information Architecture and Redesign
eXpress is a corporate messenger, video conferencing platform and super app used at large enterprises including Nornickel and Rosatom. We were redesigning the web client, and settings came with it. Rather than restyle the web section on its own, I audited settings across all three platforms, built one information structure they could share, and shipped it on web first.
- 3M+
- Users on the platform today
- Web · iOS · Android
- Platforms audited
- Audit → IA → UI → review
- Owned
Product Designer. I owned the section end to end, from the initial audit to the final implementation review, and agreed the visual language with the design lead.
problem
Support kept hearing the same thing: people could not find settings they knew were there. Nothing was missing. It was somewhere other than expected, called something else, or buried. Finding anything meant digging.
Three platforms, three structures
The same setting sat in a different place, under a different name, depending on where you opened it.
Features buried
Things people used regularly sat several levels down, with nothing in the structure to hint where.
No room to grow
There was no logic to slot a new setting into, so every addition made the section harder to navigate.
What the audit showed
I walked every platform (web, iOS and Android) and documented every settings item: structure, naming, reachability, behaviour. Put side by side, the three had visibly drifted apart.
research
Competitive analysis
I studied how Telegram, Discord, Slack and Teams organise settings: how they group, how they name sections, what they surface at the top level. Then I pulled out the patterns that work for products with a comparable audience.
All four resolve to the same shape on desktop: a standing rail of sections on the left, content on the right, and a short list of domain groups (notifications, appearance, privacy, data, devices), with the account kept apart from them. None of them makes you drill into a list and come back out. That is the pattern eXpress users already carry in from every other tool on their desk, and it is what the new structure had to meet.
design
Designing the new information architecture
I worked it out on one map: the structure as it stood on the left, the screens themselves in the middle so every item could be checked against what it actually does, and the grouping I was proposing on the right. Where I was not sure, I marked it rather than picking quietly.
Two were still open when I took it to the team. Whether saved messages, favourites, tags and stickers should collapse into one entry: they are all things you file away, but people reach for them for different reasons. And whether data and storage belonged on the web client at all, when the browser handles most of it. I presented the map, collected feedback and revised; working this way meant technical constraints and product priorities were accounted for before any visual design started.
What settled: one entry for saved messages, tags and stickers left on their own, and storage off the web section. That gave the structure below, unified across the three platforms and grouped so a new setting has somewhere obvious to go.
UI design inside the design system
I designed the settings interface for web and desktop against the design system being rolled out at the time. Some of its components weren't ready yet when I was designing, so I designed against the future standard and agreed components with the system team in parallel. That cost two rounds of revisions once the final components landed.
handoff
Handoff and design review
I prepared documentation and specs for the engineers (annotated layouts, state descriptions, edge cases) and ran a sync with development before implementation started. After the build I reviewed the result against the layouts, logged the discrepancies and agreed the fixes. That is what kept the shipped product faithful to the design.
results
- The shared structure shipped on web: settings people actually use moved up to the first level, and the naming is the one the other platforms can adopt.
- One architecture to extend instead of three to reconcile. A new setting now has an obvious place to go.
- The section became a full part of the redesigned product, running on the design system.
The lasting piece is not the web section but the structure underneath it: one map the other platforms can move onto, so the product stops drifting apart every time a setting is added.