Preserving continuity across devices.
Passengers move between the vehicle’s built-in displays and their own smartphone — adjusting comfort from the rear-seat screen, then reaching for a companion app on their phone. The device changes. The product does not.
At a glance
- Client
- Mercedes-Benz — MBUX Companion App, with the UX Design & HMI team
- Role
- UX research, product strategy, interaction design. I designed both interaction concepts, designed the study, and ran it.
- Context
- An independent research study — a Master thesis in collaboration with Mercedes-Benz. Which is exactly why the answer held: no stakeholder owned either concept.
- The problem
- Familiarity could come from the platform (iOS and Android conventions) or from the product (MBUX logic). Both credible — and undecidable on paper.
- The decision
- Stop debating. Build both interaction models, put them in front of 32 people on smartphones and tablets, and measure behaviour — not preference.
- The outcome
- On smartphones, 75% of interaction patterns favoured the MBUX-adapted model. The direction was chosen from evidence, and the continuity approach is reusable beyond this one screen.
01 — The challenge
The question wasn’t how to design another mobile app.
Most mobile applications follow the conventions of iOS or Android. That usually reduces friction — users already understand their device. But companion apps occupy a different role: passengers don’t experience them as independent products. They move between the vehicle interface and the smartphone as part of one continuous experience. Following platform conventions could make the app feel familiar as a mobile application — while forcing users to relearn interactions they already understood inside the vehicle.
Should familiarity come from the platform, or from the product itself?
02 — Testing the assumption
Two ways to create familiarity, turned into a testable hypothesis.
Rather than debating design preferences, I built both. The objective wasn’t to compare visual designs — it was to understand which interaction model people naturally relied on. Same screen, same task, two answers:
Pick a program, then flip the concept — your selection follows.
The interface followed established mobile interaction patterns. The assumption: familiar mobile conventions reduce cognitive effort.
The interface translated the interaction principles of MBUX to mobile. The assumption: knowledge acquired inside the vehicle transfers across devices.
Program
Intensity
Level — 3 of 5
Intensity
Level 3
The same screen, expressed two ways — familiar as a phone app, or as an extension of the vehicle.
03 — Research
Studying behaviour, not preference.
Thirty-two participants completed identical comfort-related tasks using both concepts, across smartphones and tablets. Instead of asking which interface they liked better, the study observed how they actually behaved: how confidently they completed tasks, whether they could predict system behaviour, and how easily previous knowledge transferred between devices.
Share of interaction patterns favouring each model, by device.
On smartphones, 75% of interaction patterns favoured the product-first (MBUX-adapted) model. On the smaller screen, passengers reused what they already knew from the vehicle.
N = 32 · identical comfort tasks on smartphone and tablet · behavioural measures, not stated preference
04 — The insight
Participants treated the app as a continuation of the vehicle.
They didn’t approach the companion app as a separate mobile application. Rather than relying on familiar smartphone conventions, they naturally reused the interaction principles they had already learned inside MBUX. The most valuable finding wasn’t about interface design — it was about continuity. When people move between touchpoints of the same product, they expect their knowledge to move with them.
Consistency isn’t primarily visual. It’s cognitive.
05 — Translating the insight
Not recreating the interface — preserving the logic.
Phones require different layouts, navigation patterns and ergonomics. The interaction language of MBUX was translated to fit the constraints of mobile, while preserving the logic users already understood. The interface adapted to the device without asking users to learn the product twice. That’s the point of the demo above: change the program, flip the concept — the task travels with you.
06 — Outcome
Behaviour, not preference, set the direction.
Watching how people actually worked replaced a design debate with a measured answer.
- A companion-app direction chosen from evidence, not design opinion.
- A mobile design language that carries MBUX logic to iOS and Android.
- A continuity approach the HMI team can reuse beyond this screen.
Reflection
I used to think continuity meant making every screen look and work the same.
It doesn’t. Continuity lives in the mental model people carry from one screen to the next. Protect that, and the interface is free to change.
Two credible directions, and no way to choose on paper?
That’s the situation this study was built for — and it’s the work I do for product teams today. I build both directions as working prototypes and put them in front of real users. Reactions replace opinions, before engineering commits. Usually 2–3 weeks from disagreement to evidence.