5
Semi-structured interviews
Portfolio
MSc project · Kingston University London · May 2026
PRE-PERFORMANCE COMPANION
A mobile companion for the three minutes before something important. Built to support readiness, not to cure nervousness.
−3.25 ↘
Nervousness
Average drop (1–10 scale, n=4)
−4.5
↘
↘
Mental clutter
Average drop
+4.25 ↗
Clarity
Average gain
+1.75 ↗
Readiness
Average gain
02 -
The problem
The problem
THREE MINUTES BEFORE SOMETHING IMPORTANT, THE BRAIN WORKS AGAINST YOU.
When users are minutes away from a presentation, interview, or high-pressure meeting, their cognitive capacity drops. Repetitive thoughts pile up, mental clutter builds, and current calming tools either take too long or demand too much attention to be useful in that moment.
I don't need a full solution. I just need something that helps me stop spiralling right before I begin.
Meditation apps are too long. Breathing tutorials demand focus. The phone is the distraction, not the fix.
03 -
The question
The question
THREE MINUTES. NO ONBOARDING. NO LONG FLOWS. NO INSTRUCTIONS TO READ. WHAT DOES SUPPORT LOOK LIKE AT THAT SCALE?
04 -
Research
Research
FIVE INTERVIEWS. SIX CLUSTERS. THREE HMWs.
6
Affinity clusters identified
3
HMW statements synthesized
Stress before performance was consistently described not as panic, but as clutter — repetitive thoughts, mental noise, loss of focus. Coping behaviour varied widely: some users wanted distraction, others wanted to offload thoughts, others wanted a passive reset.
HMW statements
HMW 1
How might we help users with performance anxiety or panic?
HMW 2
How might we help users get back into focus?
HMW 3
How might we distract users for a few seconds to break overthinking?
05 -
Co-design
Co-design
ONE 45-MINUTE SESSION. THREE FEATURE DIRECTIONS.
A co-design session with five UX design students. At first, everyone proposed an idea to each statement - then they voted for the best idea. Crazy 8 sketches per HMW idea winner, tied votes resolved with a runoff.
Selected feature directions
| HMW | Selected direction |
|---|---|
| Anxiety / panic | 30-second balance game |
| Back into focus | Focus mode blocking distractions |
| Break overthinking | Thought offloading |
A fourth mode — guided breathing — was added later based on secondary research on
acute physiological stress.
06 -
Design decisions
Design decisions
FOUR CHECKPOINTS.
01
Multi-mode, not one flow
Stress is not experienced uniformly. Built four targeted modes instead of one universal grounding sequence.
02
One action per mode
Tilt for balance, speak to offload, watch the fill, and press for breath. No menus, no scoring, no failure states.
03
Emotional tone as usability
Calm visuals and motion are not decoration. In a nervous state, tone determines whether the tool feels usable or adds load.
04
Science where method didn't reach
Co-design produced three modes. Secondary research on acute stress added the fourth: guided breathing.
07 -
The prototype
The prototype
LOW-FI TO MID-FI. TWO STAGES, TWO QUESTIONS.
Low-Fi tested whether the structure made sense. Mid-Fi tested whether the interaction felt grounding.
Iteration deltas
| Low-Fi finding | Mid-Fi change |
|---|---|
| Distraction game purpose unclear | Wording and description refined |
| Thought offloading not understandable as static | Example words and visual bubbles added |
| Static screens couldn't convey emotional tone | Visual polish moved closer to moodboard |
| Activities needed to feel real | Functional interactions added (sensor, mic, touch) |
Artefacts
Low-Fi and mid-Fi screens
Home screen
Balance game
Thought offloading
Focus fill
Breathing
08 -
Validation
Validation
FOUR STATES. FOUR USERS. PRE AND POST.
Each participant rated nervousness, mental clutter, clarity, and readiness on a 1–10 scale before and after using the mid-fi prototype.
−3.25 ↘
Nervousness
7.6 → 4.4 avg
−4.5
↘
↘
Mental clutter
9.5 → 5.0 avg
+4.25 ↗
Clarity
4.0 → 8.25 avg
+1.75 ↗
Readiness
6.25 → 8.0 avg
Three states shifted meaningfully. Readiness moved less — likely because feeling "ready" depends on factors outside the app (preparation, context, stakes).
Small sample. Scenario-based. Early indication, not clinical evidence.
09 -
What broke
What broke
THREE FAILURES THE TEST SURFACED.
01
Offloading crashed
Microphone handling failed mid-session. Error handling was thin. The feature couldn't be evaluated in that test.
02
Sensor data dropped
Balance game ball stopped responding on some devices. No calibration, no fallback state.
03
Breathing was unclear on first use
Users liked the press-to-exhale, release-to-inhale interaction once they understood it. They didn't understand it on first try.
For a tool meant to be used in stressful moments, technical friction undermines the entire premise.
10 -
Technical
Technical
BUILT IN FLUTTER. TESTED ON DEVICE.
Mid-Fi prototype implemented in Flutter, deployed to mobile. Accelerometer drives the balance game. Microphone captures offloading; touch input drives the breathing cycle.
Each mode is independently testable, so individual support modes can be refined or replaced without rebuilding the system.
11 -
Next steps
Next steps
WHERE THIS PROJECT GOES NEXT.
- 01Stabilise the microphone and sensor layers — reliability is non-negotiable for stress tools.
- 02Add first-use cues for the breathing and balance modes.
- 03Test in real pre-performance contexts, not simulated ones.
- 04Compare which mode helps which user state — not all stress is the same stress.
000 · Back to portfolio
Moritz Wallbrecher→