“Rich enough that the experts recognized their world. Simple enough that a colleague with no energy background actually wanted to explore.”
The problem
Swissgrid runs Switzerland's electricity transmission grid, and its control room is one of the most restricted spaces in the company: of ~1,000 employees, only a handful are cleared to work inside it. The goal was to give everyone else a window into that room. Security capped what we could show (delayed data, nothing infrastructure-critical), but access to the actual topics was broad. The whole design problem lived in one tension. Too dumbed-down and the experts disengage; too faithful to the complexity and everyone else bounces. My co-lead and I also started as outsiders to the energy domain, which became part of the method rather than an obstacle to it.
Approach
- 1
Unpacked the experts' "wall of wishes" (up to ten equally-urgent topics) into one question: what should everyone see first? Then one requirement at a time (a map, interactive points, a time slider, three always-on KPIs, detail on click).
- 2
Built the first draft as a faithful, literal mirror of exactly what the experts asked for. Not a solution yet, but a shared thing to react to.
- 3
Re-prioritized from the user's seat, not the expert's: made the map the surface you explore, layered the KPIs behind points of interest, scoped the slider to the topic at hand.
- 4
Let the core rule surface from that work: nothing goes deeper than three layers. The cap stopped newcomers drowning and forced us to distill each topic to its essence; that distillation is what won the experts over as respect for the domain, not a dumb-down.
The draft they asked for, then the one they kept
The first draft was a faithful mirror of the brief: three always-on KPIs across the top, a detail pane beside the map of equal weight, a slider spanning everything. Nothing is missing from it. That was the problem. Re-prioritizing left one surface to explore, moved the KPIs into a column, and let the detail share that column rather than hold ground of its own.
My contribution
- Owned the information architecture across every topic.
- Built the modular UI system, designed to scale to topics we couldn't yet foresee.
- Drove the re-prioritization and the three-layer structure the whole product runs on.
- Cut and edited the case-study film for the award submission.
Protected material
The first build and the one it became, side by side.
Swissgrid's control-room data is restricted, and these are real screens of it. They exist here rather than on the public page for the same reason the case study cannot show the product: the room a handful of people are cleared to enter is not one you publish screenshots of.
Open the protected version →Credentials are in my application, or ask me.
Outcome
~150
regular users a year on, of a product no one is obliged to open
~1,000
colleagues given a window into a room a handful can enter
3 layers
hard depth cap, the rule the whole product runs on
We unpacked every remaining topic the same way. By the second, the experts weren't being sold the structure. They'd walked the path and started applying the three-layer logic themselves, so the constraint became shared rather than imposed. A year after release, Swissgrid 24/7 still has ~150 regular users of an entirely optional product. Proof the balance held: rich enough for the experts, simple enough for everyone else.
