End-to-end process
B2C
0-1
2025
UX Unite has released a job market report in a PDF format for the past two years. Now, they want to create an interactive dashboard on their website to provide designers with updated insights into job market data and hope to turn more website visitors into community members.
My role
Product Designer
The team
Timeline
8 months
User impact
%
Comprehension rate
All test users interpreted the design correctly in final usability testing rounds.
Positive user feedback
Users found the design, calm, easy and useful.
Business impact
Reduced workload
Reduced operational overhead by automating reporting.
USP
A solution that sets itself apart by addressing confusing job titles.
Challenge
Designers lack relevant salary data
There is no reliable data in the UX field in Denmark and as a result, many designers are underpaid and face challenges in securing fair compensation as experienced by many of UX Unite's members. The company struggles to find the workforce to generate yearly reports and chasing individuals to fill out salary surveys.
By addressing this gap, designers will be empowered with data that supports fair negotiations and helps them advocate for themselves confidently.
"I used glassdoor but it was useless"
Quote from interview
Survey
Skills say more than job titles
Before interviews, I wanted to identify broader patterns with surveys. In open-ended responses, 25% noted that industry relevance was significant to them. They ranked the "most valuable data" as: 1) years of experience, 2) skills, and 3) job titles. They also reported an average confidence level of 2.7 out of 5 in negotiating, highlighting the need for support.
The survey indicated that job titles alone lack sufficient context for designers.
Interviews
Inconsistent job titles harm salary transparency
"I would love to know the main responsibilities so I can get an impression of(...) could I even compare myself?"
Quote from interview
Design
Turning constraints into direction
We were on the same page about which data to include, but the dashboard needed to be "done Yesterday". Since "responsibilities" and "industry" were new data points that we hadn't collected yet, they had to be set aside for now. Instead, I emphasized the data that was already available while making room for the new data points to be implemented smoothly when they would be ready.
To give as much relevance as possible, we created filters on "job category" and "years of experience".
First Hi-Fi wireframe of the dashboard
Making data easy to understand, fast.
I iterated and refined the cards through collaboration with the team, focusing on making the data easy to understand. Our data analyst had concerns about the interpretation of the pension in percentages stacked together with the base salary.
After experimenting with variations, I opted for vertical bars with pension in a separate column, which allows for a clear and quick overview, as well as better accessibility.
3 variations of the same salary data
V1
Stacked bar giving relative overview of the average salary
V2
Simple bars with hover view of pension and base salary
V3
Immediate overview of base salary + pension.
Reality checking design with actual data
To test how the design would perform with real data, I experimented with our dataset, revealing that our filtering was heavily fragmenting the data. I proposed merging some job categories, since we had learned that designers don't focus heavily on them. There was concern that some users might not feel represented by the titles.
Instead, we allowed designers to choose whether to combine the data and achieve meaningful results with multi-select options.
The bell curve with filters "UX Designer" + "1 Year"
Giving context to combined data categories
The multi-select functionality created a new challenge – how do we make sure designers understand the results of combined categories? I created a data pool section giving designers insight into the composition of the filtered data. I decided to add responsibility data here, not completely sure how to shape it.
First version of card showing information on filtered
Usability testing
The bell curve isn't ringing any bells
For the usability testing, I was curious about designers' reactions to the multi-select option in the filters and their understanding of the bell curve. The bell curve confused all except one designer, and many got confused or misinterpreted the chart. The data pool had mixed results; some didn’t notice it, while others struggled to figure out what to expect before opening the card.
The majority were happy using the multi-select option, while a few didn't notice it as an option.
Everyone understood this
Iteration
Simplifying salary range achieving 100% clarity
While I considered adding informative labels to the bell curve, I changed it completely. Instead, I created one simple bar that would show distribution, highlighting the 75% range, median, and lowest + highest salaries, giving the users all the information they were looking for, making the assisting cards obsolete.
I also added the data number to the data pool information, giving designers an idea of the validity of the results.
Retrospective
Knowing when to listen and when to lead
It's always a delicate dance finding common ground when people have differing views. I always want to give room for others' views, but I need to back my own views more. Often it's not the best idea, but the one that sounds most convincing that gets picked.
Visuals aid communication
While my research revealed clear user needs, I experienced resistance. I saw how visual presentations illustrating the problem/consequences made the biggest impression. It's worth the extra prep time to really think through how to make it click in people's heads.
