ALL WORK 6 MIN READ
Research · Testing

AMG Dynamic Select

AB testing three UI variants of the AMG driving mode menu under real conditions

Role UX Researcher — planning, moderation, synthesis, recommendations
Timeline 2023 – 2025
Team Mercedes-AMG HMI Team
Status Completed

AMG Dynamic Select: the most used control in the car

AMG Dynamic Select is the driving mode selector — the control that determines whether you’re in Comfort, Sport, Sport+ or Race. For AMG drivers, it’s not a settings menu; it’s an active driving tool, used multiple times per journey.

The HMI team had developed three distinct UI concepts for the next-generation selector. Each had internal advocates. Nobody had user data. My brief: run a structured comparison study and deliver a clear recommendation backed by evidence.

An 8-step research process

The study followed a structured eight-step process designed to minimise bias at every stage — from how stimuli were prepared to how results were synthesised before the team saw them.

01 Kick-off — Align on research questions and success criteria with the HMI team

02 Preparation — Stimuli design, test protocol, scenario scripting

03 User Management — Recruiting, scheduling, consent and briefing

04 Drive & Test — In-vehicle sessions on a defined test route

05 Usability Testing — Standardised questionnaires (AttrakDiff, SUS)

06 Evaluation — Affinity mapping, synthesis, findings without the team

07 Presentation — Structured readout with ranked recommendations

08 Revise — Design feedback loop based on stakeholder questions

Project plan — 8-step research process from kick-off to concept revision

8
Process phases from brief to recommendation
6
Weeks from kick-off to final presentation
17
Participants recruited across AMG driver profiles

Horizontal Picker vs Vertical Picker vs Slide Control

Horizontal Picker

Horizontal Picker

Winner

Modes in a horizontal row. Selection via lateral scroll or direct tap. Current mode centred and highlighted.

Maps to natural left-right mental model for sequential modes
Current state always visible in peripheral vision
Consistent with driver eye movement patterns
Requires wider display real estate
Vertical Picker

Vertical Picker

Modes stacked vertically. Scroll up/down to select. Similar to mobile list pickers.

Familiar interaction from mobile context
Works on narrow displays
Vertical scroll conflicts with other in-dash gestures
Requires looking away from road longer
Less intuitive mode hierarchy mapping
Slide Control

Slide Control

A single slider spanning all modes from left to right. Drag to select, snap to mode position.

Analogue feel — matches continuous spectrum of modes
Single gesture for any transition
Precise targeting difficult while driving
No clear visual state for current mode
Highest error rate in all sessions
STUDY RESULT Horizontal Picker — most preferred in study Faster task times, highest SUS score, and consistent first-choice preference across all participant segments.

17 participants. Real vehicle. Real conditions.

All sessions were conducted in the actual test vehicle on a defined route combining urban, motorway, and winding road segments. Participants interacted with each variant in counterbalanced order to control for learning effects.

Study session — participant evaluating Dynamic Select variants on the defined test route

All three variants were built to the same fidelity in ProtoPie and delivered via the same in-car simulation setup. Visual design was held constant — only the interaction pattern varied — to isolate the effect of UI structure from visual preference.

Horizontal Picker wins — for clear reasons

Post-study debrief — participants explaining their preference for the horizontal layout

71%
First preference for Horizontal Picker after all three variants
4.3/5
Average SUS score vs 3.4 and 3.1 for alternatives
2.1s
Average task time vs 3.8s and 4.4s for alternatives

The Horizontal Picker outperformed on every metric that matters for in-car interaction: task speed, error rate, subjective usability, and stated preference. Participants consistently described the horizontal layout as “obvious” and “already knowing where things were.”

The recommendation was accepted. The design team used the findings to resolve a months-long internal debate, and the Horizontal Picker moved forward into the next engineering phase.

Key Insight

The best outcome of a user study isn't a surprising result — it's a clear one. Clarity that ends debate and moves a team forward.

Research as a decision-making tool

What struck me most about this project was how much the team needed permission to decide. Three smart people had been arguing about three good options for weeks. The study didn’t reveal anything they couldn’t have guessed — but it gave them something to stand behind.

That’s what research does at its best: it transforms preference into evidence, and evidence into momentum.

Next Case

UX Maturity at AMG