60-second story
From a plausible concept to a reliable decision logic
Problem
An agency had developed 3–5 VOC display concepts from desk research. It was still open how people would understand them with a real device over several days.
Approach
Two diary studies, a randomized concept test (N = 259) and five interviews connected longitudinal use, preference and explanation.
Key insight
There was no single solution. Traffic lights enabled quick orientation, trend lines added context at a higher interpretation effort.
Decision
Both formats needed value, context and action. I translated the evidence into a recommendation catalog: when traffic light, when trend, which color, what to add and which UX/UI conditions Sales could explain to customers.
Outcome
A decision logic instead of a ranking. Subsequent implementation and customer adoption were outside my scope.
The concepts were plausible. What was missing was the reality check in daily use.
Before I joined, an external agency had conducted extensive desk research and design strategies. The presentation was well-founded and delivered several plausible display ideas – but these had not yet been tested with users and a real device over multiple days.
The Sensirion team therefore commissioned me to take the concepts developed by the designer from the strategy level into real-world use. My brief was to empirically test their strengths, limitations and conditions of use – not to confirm a pre-assumed solution.
STARTING POINT
3–5 plausible display concepts from desk research and design strategies.
CHALLENGE
Static concepts did not show how changing VOC values would be understood at home.
PROJECT GOAL
Empirically investigate daily interaction, needs, pain points and format preferences.
DECISION QUESTION
Build evidence for when each concept should be recommended – including UX/UI conditions.
Research Lead: Diana
Decision ownership: UX/UI Lead + Product Managers + Engineers
Two plausible design hypotheses, that weren’t yet validated solutions
vs
Compared display directions: traffic-light status and trend line (Two Variables: color and shape)
HYPOTHESIS A
Traffic-light status
Developed by Anna Maria: quick recognition of the current status through familiar color signals.
HYPOTHESIS B
Trend line / spline chart
Developed by Anna Maria: more temporal context by showing how VOC values change over time.
Each phase addressed a different uncertainty – from daily comprehension to recommendation
I sequenced methods so that each phase answered a different question. The diary studies were recruited through an external market research partner and incentivized due to the high commitment required. A randomized concept test then compared four value formats; five 1:1 interviews explained the reasons behind preferences.
Approach at a glance
Each method reduced a different uncertainty. Unresolved questions were deliberately carried into the next phase instead of forcing an early winner.
01
Observe in context 1
Diary Study 1, n = 9, 7 days
Test traffic-light status and color schemes in natural use; identify ambiguity and missing context.
02
Compare formats
Concept Test, N = 259, randomized
Compare four value representations and retain ambiguity instead of declaring a survey winner.
03
Explain the why
5 × 1:1 interviews
Explain preferences, expectations and misinterpretations behind the quantitative pattern.
04
Observe in context 2
Diary Study 2, n = 12, 6 days
Test trend-line display, value formats and color schemes again in everyday use.
05
Translate evidence
Cross-method synthesis
Turn trade-offs into conditions for UX/UI and Sales recommendations.
Decision rule: preference data informed the next field study; it was not treated as the final design decision.
01
Diary study 1
N = 9; 7 days, externally recruited
Evaluation of the traffic-light concept and color schemes in participants’ natural usage context.
02
Concept test + interviews
N = 259; randomized order, online
Compare four value representations and understand the reasons behind preferences and misinterpretations through five 1:1 interviews.
03
Diary study 2
N = 12; 6 days, externally recruited
Evaluation of the trend-line concept, alternative color schemes and value formats in daily use.
Diary Study 1, Research artifact and synthesis
The biggest risk was not red. It was ambiguity.
ASSUMPTION
Red in the traffic-light concept could broadly trigger anxiety or overreactions.
EVIDENCE
This broad assumption was not confirmed in Diary Study 1. Participants generally appreciated the clarity of familiar status signals; however, later interviews still revealed individual emotional associations with red.
IMPLICATION
Color could support quick classification, but could not carry meaning alone. Users needed explanations for fluctuations, reference points and guidance on what to do next.
Research pivot: Instead of continuing to ask ‘Which color is safest?’, the next phase focused on what information users need beyond the visual signal to confidently interpret VOC
Users wanted a visible value, but there was no single ‘magic format.’
A sequential-monadic concept test with N = 259 compared four value representations in randomized order. The result was valuable precisely because it produced no artificial winner: ‘No value’ was chosen least often, while the two leading numeric formats were close together.
34%
Scale
x/500
31%
Percent
%
24%
Index
visuelles / Index-Format
12%
No value
lowest preference
WHAT WAS CONFIRMED
Concrete values were perceived as more transparent
‘No value’ was chosen least often. The five interviews additionally showed that participants expected a concrete measurement for VOC – especially because other air quality indicators on the device also displayed numbers.
THE TRADE-OFF
Familiar does not automatically mean correctly understood
Percentage values felt familiar but were interpreted by some users as an absolute health or quality score. The x/500 scale was less familiar but more neutral. Both formats therefore needed context rather than being treated as self-explanatory.
Concept Test, Setup, preference data and synthesis
Decision sequence: Both leading value representations advanced to the next real-world study. The survey preference was deliberately not treated as the final design decision.
Trend lines provided more information and increased interpretation effort
The second diary study tested the information-richer, trend-based concept over six days with N = 12 participants. This phase examined whether historical context actually helped when the display needed to work at a glance in daily life.
01
Current status needed priority
Trend lines were conceptually helpful, but small display size, contrast and color differentiation made it difficult to quickly grasp the current VOC status.
02
Values still needed context
Neither % nor x/500 was self-explanatory. Users still needed a reference frame to assess whether a value was good, acceptable or concerning.
03
Expectations for action signals were clear
11 of 12 participants expected a proactive signal – such as a notification, blinking or sound – when air quality required attention.
Evidence boundary: The research supports these usability and interpretation findings; it does not document which downstream screen variant was ultimately implemented by individual customers.
Diary Study 2, Screens and cross-method synthesis
Product context – Sensirion Indoor Air Quality Monitor: example photos from diary study participants sent to me.
The research produced no universal winner. It delivered a reliable recommendation logic for both concepts.
Both visualizations remained part of the recommendation catalog: the traffic-light concept supported quick status recognition; trend lines provided historical context but required stronger UX/UI support.
Across both, status, value, context and action had to be connected.
STATUS
Make the current status immediately recognizable.
VALUE
Show a concrete measurement instead of relying on color alone.
CONTEXT
Explain what the value and its fluctuations mean.
ACTION
Link critical states with action guidance or proactive alerts.
Recommendation logic: Traffic-light status for quick orientation; trend lines when temporal context matters and current status remains visually prioritized
From a presumed concept comparison to an evidence-based recommendation catalog for Sales
Both display directions could be recommended – under different UX/UI conditions. I synthesized the cross-method evidence into a catalog that explained when each concept was appropriate, what trade-offs existed and how Sales could justify the recommendation to B2B partners.
WHAT I DELIVERED
Evidence-based recommendation catalog
A prioritized set of conditions of use and UX/UI principles for traffic-light and trend-line concepts.
Decision rationale for Sales
Guidelines enabling Sales to justify when which VOC concept suited a customer context and which trade-offs needed consideration.
Cross-method synthesis
A consistent recommendation based on longitudinal use, five qualitative interviews and randomized preference data.
MY ROLE
Lead UX researcher
I owned research planning, recruiting partner management and the design, execution and analysis of the diary studies, concept test and five interviews – including synthesis, stakeholder readouts and the final recommendation catalog.
Cross-functional collaboration
I collaborated with UX/UI, Product, Engineering, Sales, Field Application Engineering and Software Development.
Decision ownership
Downstream product, design and customer decisions rested with the UX/UI Lead, Product Managers and Engineers – and were outside my scope.
The full recommendation catalog remains confidential. The abstracted decision logic is shown above.
WHAT I DO NOT CLAIM AS IMPACT
Which screen variants individual customers ultimately chose, how Sales used the catalog and what was implemented were outside my visibility. I therefore do not present these as impact.
Not every strong research outcome ends in a single solution
01
Desk research delivers hypotheses, not real-world evidence
The agency work was strategically plausible. Only longitudinal use with a real device showed which ideas remained comprehensible even as values changed.
02
Sequence methods by uncertainty
The concept test revealed preferences, interviews explained expectations and diary studies made long-term comprehension and behavior visible. No single method could have supported the decision alone.
03
A catalog can be more valuable than one clear winner
There was no single solution. The strategic value lay in translating trade-offs into reliable selection rules: when traffic-light, when trend and which UX/UI conditions apply in each case.
Recommended next step: Validate the recommendation logic with broader and culturally diverse user groups, test onboarding and thresholds and measure actual catalog adoption by Sales and Product. This downstream impact was outside my scope.
