Join Us needed young people for Social Booster, its government-funded programme against loneliness. TikTok ads reached 200,000+ people, yet nobody new signed up. I compared its two recruitment routes, researched who to reach, and redesigned the sign-up page: from 0 to 20 sign-ups.
Old page → new page
2025
In short
Each part keeps its colour further down the page, so you can jump straight to the one you need.
01 · The brief
Join Us asked me to recruit young people who want to make an impact, without their own experience of loneliness. Behind the question sat a hard target.
The programme works well for young people who have felt lonely themselves. We also want to reach young people who want to make an impact without that experience. Our current channels don’t reach them.
Your assignment
What was at stake
Dashed line: below 85 finishers the grant goes down. Solid line: the target.
Source: MDT grant rules (VWS) and the coordinator. The 85 is my reading: 85% of the 100 planned places.
02 · Process
My project plan, in the three phases of the Behavioural Design Method. Each dashed line is a step I went back to, when the scope changed or the interviews showed something new.
Part 1 of 3
Short answer
At the page. Open sign-up is the right route and the ad was in range: people clicked, then stopped.
03 · Recruitment data
Join Us recruits through schools and through open sign-ups. I mapped both routes and measured how long it takes to find one Social Booster.
Email a school, then wait for a whole class.
≈ 1 year of emails back and forth
Ads and the website send people to one page. From there, you email the coordinator.
04 · Literature
I checked both routes against research on volunteer motivation and team performance.
| What research says matters | Route A · school | Route B · open |
|---|---|---|
| Own choice Choosing your own role feeds intrinsic motivation. Deci & Ryan (self-determination theory) · Hackman & Oldham | The school picks | They pick |
| Mixed backgrounds Mixed teams come up with more usable ideas, up to a point. Ou, Goldschmidt & Erez (2023) · Van Zijl et al. (2023) | One study | Any study |
| Independent of schools A group can start without a year of emails first. Recruitment flow, Join Us | Needs a yes | Always open |
| People who joined What each route delivered so far. Join Us data | ≈ 72 a year | 0 new |
So I focused on open sign-up
It wins on choice, mix and independence. The only thing it lost on was volume, and the data said why: people clicked, then stopped.
05 · Benchmarks
Before touching the design, I looked for recent ad-to-sign-up funnels with published results.
Girlguiding UK
Charity · volunteers
Reached people already searching, and sent every click to one page with one action.
9 → 178 a month
volunteer sign-ups, 2,139 in a year
Goodcall
Software · sign-ups
Ads that looked like a creator’s post, not an ad, with tight targeting.
$185 → $7
cost per sign-up, 96% lower
Nolan Bros
E-commerce · sales
The page repeated the ad’s message and images. Retargeting brought visitors back.
1.2% → 3.4%
clicked the ad → bought after the click
Join Us, before
Non-profit · Social Boosters
People clicked. The page lost them: 758 words and no way to sign up on it.
0.64% → 0%
clicked the ad → signed up
Join Us next to the benchmarks
What I took from it
The click rate was within range of a strong commercial campaign. The drop happened after the click, so that is where the work went.
Part 2 of 3
Short answer
Creative students who want a real case for their portfolio. The page asked too much of them, and the brand didn’t feel like theirs.
06 · Target group
Social Boosters meet at the office in ’s-Hertogenbosch. Research on travel time set the limit, and CBS data put the focus on cities.
Who I researched
Within 30 min by train of ’s-Hertogenbosch
Within 30 minutes of the office
How far volunteers will travel
07 · Interviews
Participants clicked a TikTok ad and used the page on their phone while I watched. The line shows where they dropped off, and the psychology behind each step.
Four of my five interviewees looked like the same person. Her reason to join isn’t loneliness, it’s her portfolio.
“Working on my future matters to me.”
Job to be done
Build her portfolio, so she has a better chance at a job in visual design.
Where to reach her
Persona coded from my interviews. Name and details are a composite.
Making the research usable
Not everyone on the team knew qualitative coding. So I turned the coded interviews into a small tool: it opens on the main persona, and each dream outcome shows the forces behind it.
08 · Survey
Next to the usability barriers, interviewees judged the brand itself: would they want to be seen joining this? Before redesigning, I checked that in a larger group with a survey on brand coolness.
First impressions
Now that you’ve seen and used the website, what’s your first impression?
My question, after the site test
Answers from the interviews, translated from Dutch. The tags are my coding against the brand coolness scale.
“I’d fall asleep, a hundred percent.”Low on exciting
“Am I even allowed to be creative here?”Low on subcultural
“Looks like I’d have to put my creative energy in airplane mode.”Low on exciting
“Giving corporate.”Low on subcultural
“Reminds me of when data analytics came to the careers fair to recruit students.”Low on aesthetic appeal
“Did your web designer used to be a tax adviser?”Low on aesthetic appeal
Brand coolness of Join Us
What it means for recruiting
For a creative student, a programme is also something to show. Join Us passed on meaning and failed on look and feel. So the redesign had to raise desirability without losing the mission.
Part 3 of 3
Short answer
Three changes for one target group: sign up on the page, content for creative students and a brand of its own. Sign-ups went from 0 to 20.
09 · Strategy
I scored every opportunity on impact, effort and confidence. Normally I test one at a time. With one intake a year, I built the three that serve one target group together: creative students.
My ICE scores
| Opportunity | Impact | Effort | Confidence |
|---|---|---|---|
| Shorter time to sign up On the live page | 6 | 2 | 10 |
| Content for one target group On the live page | 10 | 5 | 9 |
| New visual identity On the live page | 9 | 6 | 10 |
| Personalised ad targeting Not in this round | 8 | 8 | 9 |
| Better TikTok ad performance Not in this round | 8 | 10 | 5 |
Scores 1–10. Effort: higher means more work. Bubble size shows confidence; filled bubbles were built.
10 · Design
The old page asked people to read 758 words and send an email to sign up. Here is the top of both pages. The full new page is available on request.
11 · Iterations
I matched each barrier from the interviews to one change on the page, and went from lo-fi to hi-fi in eight versions. Each pin is one barrier and its fix.
v0.1 · lo-fi Key facts and a waitlist field in the first screen. 
v0.5 · mid-fi Built within the CRM’s limits, so it could be tested fast. 
v0.8 · hi-fi Roles, coaches, reviews and FAQ, in the Join Us style. Home v0.45 · final design A Social Booster brand of its own, made to score higher on exciting, aesthetic and subcultural. This is the new page above. 12 · Learnings
Interest wasn’t the problem: 1,309 people clicked. The route only broke at the last step, an email nobody wrote.
A good cause isn’t enough to recruit. Creative students judged the programme like a brand they would put their name to.
Research helps only if the team can use it. A persona dashboard made my interview codes usable for the comms team.
Result
From 0 to 20 sign-ups. The cost per sign-up is now the total spend divided by 20, where before nobody signed up at all.
More work
Garden firms lose hours on requests that never fit. I designed and built a six-question request flow that tells them the budget, the garden and the wishes before the first call.
A visual identity that puts the nominees in the spotlight: logo, type, colour and a poster system, captured in a brand book.
Available for freelance work now. Remote, or on-site anywhere in the Netherlands.
Comments
2