feike.studio
Work Process About CV Contact

Case study — Vattenfall · 2024

Making a household's
energy system steerable.

Dynamic tariffs, solar, a battery, a wallbox, a heat pump. How do you turn a decentralised energy system into something a household can read at a glance and act on?

Client
Vattenfall
Role
Product Designer & Product Owner
Year
2024
Platform
iOS · Android · Web
Scope
Product · UX · UI · Brand · PO
Product
prosumer-app.de
Three phone screens from the Prosumer App: a morning dashboard with self-sufficiency, current electricity price and price development, beside live energy flow and device views.

One product core, three platforms, and a shell every utility can brand as their own. The home screen answers the two questions a prosumer actually has: what is my system doing right now, and is this a good moment to use power?

Interface in German · click a screen to enlarge

In short

A white-label portal for the energy transition

The Prosumer App lets energy providers offer dynamic tariffs, smart meter integration, live monitoring and home energy management under their own brand — on Android, iOS and in the browser.

Households are no longer just consumers. They produce with solar, store in batteries, charge cars and heat with pumps — and with dynamic tariffs the price of a kilowatt hour changes through the day. The value only materialises if someone can see the system and act on it.

The existing tools did not make that possible: readings without interpretation, device apps that ignore each other, tariffs explained in tables. Vattenfall needed a customisable product that works for their own customers and for other utilities.

Energy data is abundant. The scarce thing is a moment where a person knows what to do with it.

The challenge

Three tensions in one app

Live data vs. Calm
Flows, prices and forecasts change constantly. An interface that reacts to every value becomes a monitoring station nobody opens twice.
Automation vs. Understanding
Surplus charging and load shifting only earn trust if a person can see afterwards why the wallbox did what it did.
White label vs. Product identity
Every utility brands the app as theirs. The design had to survive foreign colours, logos and tone without losing its own structure.
Three platforms vs. One team
iOS, Android and web had to feel like one product while respecting what each platform does natively.

None of these are visual problems. They are decisions about what the app claims, what it decides on its own, and what it hands back to the person.

My contribution

Design and product ownership in one role

I shaped the product and its UX across all three platforms, designed the interface and the brandable visual system, and took on product ownership: priorities, stakeholder alignment and delivery.

The double role mattered here. Deciding what a tariff simulation may promise is a product decision, a legal one and a design one at the same time — and it is faster to resolve when the person specifying it also has to draw it.

I also owned the handover: brand-consistent, developer-ready specs for a team building the same product three times, plus the theming rules a new utility gets configured with.

Approach

Three decisions that shaped the app

01

One picture of the whole household

The live view puts grid, solar, battery, wallbox and heat pump into a single flow diagram instead of five separate device screens. Every value is attached to the thing it belongs to, so the picture reads as a house, not a dashboard.

Self-sufficiency became the headline figure — one number, with a comparison to comparable installations, for current, monthly and yearly. It gives the system a score a household can actually improve, which no single kilowatt reading does.

Live — the flow between grid, solar, battery, wallbox and heat pump in one diagram, with self-sufficiency as the figure to improve.
Wallbox — the car charges on solar surplus, and the screen says so while it happens: mode, current power and the session as a chart.

02

A price is only useful as a decision

A dynamic tariff produces a number every quarter of an hour. On its own that is trivia. The tariff view therefore states the current price, classifies it — this is a cheap moment — and shows the forecast, so the decision to run the washing machine now or in three hours is possible without reading a table.

Notifications extend the same logic to the time the app is closed: an alert when the price is heading below average, and one when it is heading above. Push is used for moments that change behaviour, not for status updates.

The price is stated, then judged — “very low, a good moment to use energy” — with the forecast one tap away.
Two alerts that matter — price dropping below or climbing above average — plus provider messages, kept in an inbox rather than the flow.
Smart meter data at 15-minute resolution, per device — the evidence behind every claim the app makes about a household's consumption.

03

White label without watering down

A white-label product usually degrades into a lowest common denominator. The way out was to separate what a utility configures from what the product guarantees: colour, logo, typography and tariff wording are theirs; layout, hierarchy, states and the meaning of a component are not.

That split is also what made three platforms affordable. One structure, one set of states, native behaviour where it counts — and a handover a distributed team could build from without asking what a screen was supposed to mean.

Result

Shipped, branded, rolling out

20+
utilities the white-label app is rolling out to
3
platforms from one product core: iOS, Android, web
15 min
resolution for smart meter consumption data

Real-time insight, tariff simulation and device control now live in one app that utilities ship under their own brand. It is rolling out to more than 20 utilities — the same product core, configured rather than rebuilt each time.

A judgement beats a number.

“Very low price, good moment” changed behaviour where a cents-per-kilowatt-hour figure did not.

Automation has to narrate itself.

Surplus charging is only trusted once the app shows what it decided and why, while it happens.

White label survives on the boundary.

Naming exactly what a utility may change is what keeps twenty brands from eroding one product.

One figure to improve.

Self-sufficiency gave a household a goal, which is what turned monitoring into engagement.

What I take from it

Nobody wants to manage energy. They want to know whether now is a good moment.

The app got better every time it interpreted instead of displaying. That is the same instinct I now bring to AI products: a system that has data has to take a position, and then be honest about how it got there.

Owning product and design together was the other lesson. The hard calls in this project were never “which chart” — they were what the app is allowed to promise about a household's money.