← Workerbird Nesta CareerTech Challenge · £50k grant, one of 20 funded finalists

Career Mapper

Career Mapper helped workers reflect on their working life, recognise the skills they already used, and see where those skills could take them, using real labour market data.

I founded Workerbird and led this project: the research, the design and the relationship with Nesta. Two developers built it.

Piloted with 460 union members: 71% came away feeling confident or empowered, and 86% found it very easy to use.

The Career Mapper landing page: a coral banner reading Career Mapper, find out what jobs could work for you, a job title search box, and three benefits — The Career Mapper landing page.

The problem

Our research found that many workers at risk from automation struggled to name the specific skills they used every day. They tended to think in general terms, which made it hard to picture other jobs. Some were also "reactive" job searchers who only looked for work after redundancy. Existing skills surveys were either too generic to be useful or built around employers' needs rather than workers'.

A completed remote research interview board on Miro, with sticky notes in blue, purple and green — A completed research interview with a footwear production operative. Blue: skills in their own words. Purple: ESCO skills and job descriptions. Green: digital skills.

A shop assistant's Miro board sorting essential and optional skills — A shop assistant sorting the essential (purple) and optional (blue) skills for their job.

A logistics worker's Miro board of skills used daily — A logistics worker identifying the skills they use every day.

The solution

We built two tools that could be used in either order.

Skills Assessment

Users picked their job, rated each of its skills from the ESCO database on a simple confidence scale, added their own skills, and downloaded a PDF summary.

Screen: What do you do? A job title search with suggestions such as warehouse worker and forklift operator — 1. Find your job

Screen: rating the skill check shipments from A, no knowledge, to D, very confident, with a Not relevant option — 2. Rate each skill

Screen: a checklist of general skills such as work as a team, delegate activities and troubleshoot — 3. Add general skills

Screen: Your Skills, grouping skills under very confident, confident, basic knowledge and no knowledge — 4. See your results

Job Explorer

A mobile-first, scrollable view of job matches, each compared with the user's current job on pay, hours, sector growth and automation risk. Green and red with arrow icons showed at a glance whether a job was better or worse. Each match listed skills they already had, related skills and new skills to learn.

We called this an "open algorithm". Rather than giving one answer, users could adjust filters (confidence or enjoyment, how similar to their current job, regional or national) and watch the results change. Behind this, each skill was scored on how confident the user felt and how much they enjoyed it, and jobs could be matched on exact skills or on the broader skill groups they belong to.

Screen: Your job insights, comparing hourly pay of £9.92 with a national average of £15.09, annual pay and hours worked — Compare your job with the national average

Screen: Job Explorer showing a match, Inventory Coordinator, at £9.75 an hour, £21k a year, 45 hours and high automation risk — Explore matched jobs

Screen: filters with sliders from Happy to Confident and from Similar to Different, and a choice of national or regional results — Set your own filters

“The salary comparison has allowed me to see how I can achieve comparable pay in a different job using existing skills. Also shows what skills I need to develop to secure a new job role.” Career Mapper user

How we built it

We ran two design sprints and remote research interviews on Miro with factory, retail, steel and logistics workers. We tested Figma prototypes personalised with each participant's real job and skills before launching a live pilot.

People’s own job titles often weren’t in the ESCO database. Our interviewees called themselves “clickers”, which ESCO lists as “cutting machine operator”, and “retail assistant” wasn’t listed at all. So the search suggests job titles as you type, and a short guided route helps anyone who still can’t find their job.

Testing shaped many details. "Skills audit" became "skills assessment" because "audit" sounded scary. A confidence scale replaced "beginner to expert" because users didn't relate to "expert".

Session recordings showed nobody opened the tab holding their matched skills, so we moved the skills back onto the main screen. Colour was never the only signal: arrows show better or worse too.

The site met WCAG AA, was fully keyboard navigable, and used privacy-first analytics (Matomo rather than Google Analytics).

Screen: How the job explorer works, explaining how to scroll through matches, compare averages and get further information — Clear guidance before the Job Explorer.

Rating screen version: Beginner, Learner, Teacher, Expert as a list with face icons — Beginner to expert, as a list

Rating screen version: four cards labelled Beginner, Learner, Teacher and Expert — Beginner to expert, as cards

Rating screen version: four cards with faces from No knowledge to Very confident — The confidence scale we kept

We tested a list against a grid, and a beginner-to-expert scale against a confidence scale. Most people preferred the grid, where the whole square is the button.

“I don’t like the word expert – bigging myself up. I’m good at my job – I know that.” Research participant

Testing what to build next

Before building anything for training or job vacancies, we added buttons for current vacancies and further training to job matches in the live Job Explorer and tracked how many people clicked them, to measure demand. A CV Builder was the next planned feature, building on the skills PDF users could already download. In the union pilot, the closing screens asked people whether they would like to learn about other sectors they could work in, and invited them to join a mailing list for the Job Explorer and the Workerbird app.

Job match screen with growth over ten years, skills for this job, and a Current Vacancies button — A button on each job match

Screen: We will be providing information on training and vacancies in the next phase of development. Please carry on exploring your matches and information. — What people saw when they clicked

Screen: CV Builder, explaining that after the Skills Assessment users could enter their details and download an editable CV — CV Builder, the next planned feature

Results

More than 460 members of our union partner completed the pilot.

86%found it very easy to use

71%felt confident or empowered afterwards

86%became interested in further training

60%wanted to know how their skills applied to other sectors (72% in logistics)

30%downloaded their skills summary

71%found the skills PDF useful

One interviewee's reaction on seeing their full skills list summed it up: "Wow, I do a lot!"

We ran two versions of the site: one starting with the Skills Assessment and one starting with the Job Explorer. Early data suggested people who started with the Job Explorer were more engaged. They seemed more motivated to finish the Skills Assessment once they had looked at jobs, and downloaded their skills PDF in higher numbers.

Beyond individuals

For our union partner we built a real-time dashboard of anonymised results by sector, which helped launch their Skills Campaign. The union shared the findings with its sector reps and learning and education teams.

We also added 15 custom screens with yes or no questions on literacy, numeracy and digital skills. Anyone who answered no, or rated a job skill as no or basic knowledge, got links to the union’s learning resources in their results. Each organisation could be given its own access code rather than a separate site. This piloted an enterprise model designed to keep the tool free for workers. We also designed how the Job Explorer could work inside a company, to show internal career paths or staff internal projects.

Union version start screen explaining the assessment was developed for union members, is completely anonymous, and asking for consent to store data — Built for union members: anonymous, with consent

Screen introducing the English section, 1 of 3 — English, maths and digital sections

Screen: I can easily add 2 numbers under 100 in my head, with Yes and No buttons — Simple yes or no statements

Page 1 of a completed skills report for a warehouse worker, listing skills rated very confident, confident and basic knowledge

Page 3 of the same report: English 4 out of 5, maths 5 out of 5, digital 5 out of 5, and next steps linking to the union's courses and skills academy

A real report a member downloaded at the end, anonymous, with their skills, general skills scores and next steps.

Charts from the union Skills Assessment dashboard showing skills confidence by sector — Part of the Skills Assessment dashboard built for our union partner.

← Back to Workerbird