The Market Problem.
Employers need to attract candidates, but running recruitment advertising introduces decisions about channels, content and budget. The design opportunity is to make those decisions approachable.
AI RecruitmentMarketing Platform.
Making a complex recruitment workflow easier to navigate, with AI recommendations people can review, understand and correct.
From a job description to a reviewable vacancy.
01 / Context
Background.
Personeel.com is a Dutch B2B recruitment marketing platform for employers who want to manage their own recruitment campaigns.
The gap.
Creating a campaign involves more than writing a vacancy. Employers need to choose channels, prepare content, understand costs and follow up with candidates, often without specialist marketing knowledge.
The goal.
Bring those tasks into a coherent workflow. Use AI to help with preparation while making its recommendations understandable, editable and subject to human review.
02 / Discovery
Research&Discovery
Understand the work behind the campaign before designing the workflow.
I interviewed account managers and stakeholders, benchmarked recruitment tools and used usability testing to connect workflow insights to product decisions.
Problem for Businesses.
A self-service product needs to support the work that account managers previously guided. Missing information and unclear next steps can interrupt a campaign before it is ready.
Who we designed for
Two working archetypes helped frame different needs for guidance and control.
Archetype 01The time-strapped owner
A clear next step, minimal setup and guidance that does not assume marketing expertise.
Archetype 02The proof-driven recruiter
More control over campaigns and clear information to understand and communicate performance.
Account manager interview
These conversations helped translate the existing service into product requirements: what information is collected, where preparation breaks down and which decisions still need human judgment.
Key Insights
01
Make the intake explicit
Map the information needed for a campaign, rather than treating the job description as the complete brief.
02
Bring asset preparation into the flow
Photos and channel requirements need clear prompts at the point where they become relevant.
03
Keep follow-up connected
The workflow continues after an application arrives. Candidate management belongs in the product scope.
04
Identify repeatable screening work
The manual setup behind screening highlighted an opportunity to connect vacancy requirements with configurable screening rules.
03 / Validation
UsabilityTesting
A good extraction is only useful if people can finish the workflow.
Usability testing highlighted friction in asset handling and unclear validation. I used the findings to separate implementation failures from interaction problems, so each issue could be addressed by the right owner.
Observed frictionAsset preparation
Asset preparation
Unclear validation
Design responseGuided review
Guided review
Actionable feedback
How I organised the findings +
Implementation issues
Failures such as unresponsive controls and broken uploads need reproducible steps and an engineering owner.
Interaction issues
Unexplained requirements and ambiguous recommendations need clearer guidance and a design response.
04 / Visual direction
Avisuallanguage,builttoscale.
From a supplied identity to a visual language of my own.
The interface in motion
The product in motion.
Nine product screens, composed into a continuous motion study for this case study.
The starting point / Existing identity
The identity, distilled.
Given: logo + core colors. Developed: visual direction, supporting palette and interface system.
Personeel.com / Existing wordmark
Wordmark alignment study · Existing logo
Orange#FF5B2D
Navy#1B2B4A
Supporting blue#ECF3FD
Surface#F8FAFC
Brand accents → supporting surfaces.
Typography & colors
Montserrat
Headings
ABCDEFGHIJKLMNOPQRSTUVWXYZ
0123456789
Headline AaSubheadline AaBody text Aa
Textabcdefghijklmnopqrstuvwxyz
0123456789
Four directions.
One brand.
I explored character, contrast, data and atmosphere to define how Personeel could feel beyond its existing logo and colors.
01 / Friendly AI
A character-led welcome makes the AI proposition feel approachable.
02 / Select · Option 04Less noise.
Less noise.
More presence.
Selected direction
I developed the four concepts; leadership selected option 4 for its simple, premium feel.
A single focal point.
The centered form gives the page a clear visual destination.
Brand as atmosphere.
Orange and blue move to the edges; the action retains a solid orange accent.
Depth, with restraint.
Translucency, fine texture and curved light create detail around a simple layout.
Expression at the entrance.
Clarity in the work.
The visual system takes on a functional role inside the product: distinguish information, reveal exceptions and make the next action clear.
Interface values illustrate the design; they are not measured product outcomes.
A shared design foundation.
I built a 30+ component system with Figma Variables and tokens, from visual foundations to interaction states and handoff guidance.
Color with a defined purpose.
Brand palettes, neutral surfaces and semantic status colors establish a shared visual language.
Brand · Neutrals · Status colors05 / Design
ProductDecisions
Each decision starts with a user problem. The interface makes the response tangible.
Flow & Structure
One connected journey, from creating a vacancy to managing the people who apply.
Explore each step +
Create vacancyBring the job description into the workflow.
01
AI Create Vacancy
How can an employer judge a suggestion they did not write?
Switch to compare
02 / 02
Redesigned UI
Make AI output easier to review and correct
I designed uncertainty flags, plain-language explanations and quick overrides so employers can question a recommendation and correct it. The split view brings source material and the vacancy closer to the editing task.
Connect suggestions to their source
The source document sits alongside the editor so employers can compare extracted details with the original content before confirming them.
Direct attention to fields that need review
Uncertainty flags distinguish fields that need a closer look. Plain-language explanations help users understand why a suggestion needs their judgment.
Keep the employer in control of the final content
Employers can edit or override AI-generated recommendations. AI helps prepare the vacancy; the person reviewing it makes the final decision.
02
Platform Selections
Support a channel decision without hiding the reasoning.
Switch to compare
02 / 02
Redesigned UI
Offer a starting point, keep the choice visible
The revised layout separates included channels from paid options and makes recommendations easier to scan. Explanations and overrides remain essential: the recommendation supports the employer’s decision.
Surface what is already included
Included channels are separated from paid options, making the distinction between available value and additional spend clearer.
Make selected channels explicit
Visible selection states let employers review their channel mix and change individual choices before continuing.
Explain recommendations instead of relying on a score
A recommendation needs a reason that an employer can assess. The user should be able to question it and choose a different channel.
A learning from testing: a match score can steer a choice without making it better informed. The rationale matters as much as the recommendation.
03
Content Studio
Shift the work from configuring every variation to reviewing useful options.
I explored three ways to give employers control over campaign visuals. Each iteration changed when that control appeared: from configuring everything upfront, to following a sequence, to reviewing a prepared starting point.
- 01 / Expose the controlsTemplates, formats and creative settings in one place. Flexible, but the employer had to understand the production process before seeing an ad.
- 02 / Guide the sequenceBreaking the work into steps clarified the order. It still left asset preparation and repeated creative choices with the employer.
- 03 / Curate, then refineStart with generated options. Keep brand settings, templates and individual edits available when a specific change is needed.
The shift was in when to offer control. I moved detailed configuration out of the starting path, while keeping review and correction in the employer’s hands.
Explore the iterations
03 / 03
03 / AI curator
A useful starting point, with room to override
The third direction changes the starting task from configuring an ad to judging prepared options. Brand settings, platform tabs and individual edits remain available, so employers can correct an output without managing every creative setting upfront. AI supports preparation; the employer keeps the final decision.
04
Review before Launch
A clear checkpoint before committing to a campaign.
The review page brings the creative preview and itemised costs together. Employers can check the content, distinguish included channels from paid spend and understand the commitment before continuing.
05
Beyond the launch
Keep campaigns, performance and candidate follow-up connected.
Explore the iterations
01 / 03
Vacancies
Keep campaign health visible
Campaign and channel status belong together. A platform problem needs to be visible in the overview, even when the overall campaign is active.
06 / Measurement
Whattomeasure.
Finishing a form is not the same as making a successful hire.
The measurement focus is the journey from seeing a vacancy to submitting an application and progressing as a candidate. Each stage answers a different question.
- 01
Reach the action
Can visitors find and use the application entry point?
- 02
Complete the application
Where do form or validation failures interrupt the journey?
- 03
Evaluate candidate quality
Do applications progress to useful conversations and appointments?
This section describes the evaluation framework. It does not claim a measured conversion lift.
07 / Reflection
Outcome&learning.
A connected recruitment experience, with human judgment built into the AI workflow.
My contribution spans research, interaction design and a shared component system across the recruitment journey. The work brings vacancy creation, campaigns, content, payment and candidate management into a common design framework.
What I learned.
The most useful design input often sits inside a manual workflow. Understanding that work helps identify what can be simplified, what can be supported by AI and what still needs a person’s decision.
What comes next.
Continue validating the end-to-end flow with employers, prioritise the remaining friction and distinguish implementation progress from measured user outcomes.