← Article
Brink · User Research Document
Research plan Empathy map Pain points Personas User stories Journey map Problem Hypothesis Value prop Metrics
Product Design · Professional Activation

Brink — User Research Document

The consolidated research for Brink's professional-side activation flow: the journey from sign-up to becoming "assignment-ready" and seeing a first relevant match. This is the first diamond — Discover and Define — written up as one artifact, and it is the context handed to the agent before any prototyping begins.

About this document. The interviews, numbers and quotes here are constructed — built to reflect how the Dutch independent-professional market actually behaves, and to stay internally consistent so the synthesis is fair. In a live engagement this is real work the research phase produces over ~3–4 weeks. It accompanies the Double Diamond guide and the full dossier (documents 00–14).
Part 1 — Research plan
1 · Project background

What led to this research

Brink is a two-sided talent marketplace in the Netherlands, connecting independent professionals (ZZP'ers — freelancers, contractors, interim specialists) with partner organizations for project assignments. It is evolving from a structured listings board into a modular, intelligence-driven marketplace — AI inside discovery, matching and decision-making rather than bolted on the side.

New professionals sign up — usually after seeing a relevant assignment — but a large share never become "assignment-ready" (profile complete, rate set, identity and compliance verified, visible to partners). The drop-off isn't at the front door and isn't at the very end. It's in the messy middle, where the platform is currently silent and form-driven. This research investigates that middle on the professional (consumer) side.

Source: 00-brief.md §1–2

2 · Research goals

The design problems we're trying to solve

Understand why professionals stall between sign-up and assignment-ready, in their own words — separating real blockers from assumed ones — so we can move the activation funnel without sacrificing profile quality or trust. Specifically:

  • Locate where effort is disproportionate to reward, and why people abandon there.
  • Understand the rate-setting decision and what makes it hard.
  • Learn how professionals react to compliance, and when it feels wrong.
  • Establish what would make them trust — or distrust — AI doing parts of this for them.

Source: 00-brief.md §2–3, 01-research-plan.md (goal)

3 · Research questions

What the research set out to answer

  1. What actually happens between signing up and either finishing or abandoning the profile?
  2. Where does effort feel disproportionate to reward, and why?
  3. How do professionals decide their rate — and what makes that hard?
  4. How do they react to compliance requests (KvK, BTW, ID, insurance), and when?
  5. What would make them trust (or distrust) AI doing parts of this for them?
  6. What does "this was worth it" feel like — when does activation emotionally land?

Source: 01-research-plan.md (research questions)

4 · Key performance indicators

KPIs toward the end goal

The activation funnel we are trying to move, and the single number that fuses both success events — the platform's (ready) and the user's (a match).

Funnel stepBaselineTarget
Sign up → profile started71%
Profile started → assignment-ready the leak34%≥50%
Assignment-ready → first match within 30 days58%70%
North Star — 7-day activation (net sign-up → ready + first match)~24%40%
Median time-to-ready (those who finish)9 days<1 day
Guardrails (must not degrade): profile-quality score ≥ baseline · AI-draft acceptance ≥ 70% kept with light edits · AI opt-out < 15%. The goal forbids the cheap win of flooding the marketplace with thin profiles.

Source: 00-brief.md §2 (funnel), 12-product-goal-success-metrics.md

5 · Methodology

The steps we took

  • 6 × 45-minute semi-structured interviews, remote — one unusable (audio loss), leaving 5 transcripts.
  • Think-aloud walkthrough of the current onboarding where the participant still had access.
  • Single screening criterion: independent professional in NL who created (or seriously attempted) a Brink profile in the last 6 months.
  • Analysis: affinity mapping across transcripts → recurring tensions → insights. Each insight must trace to at least two participants to count as a pattern, not an anecdote. Quotes are tagged by participant so claims stay auditable.

Source: 01-research-plan.md (method, analysis approach)

6 · Participants

Who we spoke to, and why

Sampled deliberately across the experience spectrum — not only people who churned — to avoid confirmation bias. Includes people who succeeded and one who left for a competitor.

CodePseudonymProfileOutcome on Brink
P1SanneExperienced interim PM, 6 yrs ZZP, time-poorStarted twice, abandoned both
P2MehmetNewly independent (3 months), ex-employeeFinished slowly (11 days), anxious
P3JoostVeteran contractor, 14 yrs, platform-skepticalAbandoned — "too much hand-holding"
P4FatimaMid-career UX/research freelancerFinished, actively matched — success case
P5TomNewly independent developerAbandoned, went to a competitor

Source: 01-research-plan.md (sampling)

7 · Script

What we asked participants

A semi-structured moderator guide — a spine, not a verbatim script — followed by the think-aloud protocol. Every block ties to a research question above. The full guide is in 01a — Interview & Think-Aloud Script; the spine:

1 · Warm-up & context RQ6

  • Tell me about your work as an independent.
  • What made you sign up that day?
  • What were you hoping would happen after?

2 · Narrative replay RQ1–2

  • Take me from account-created to finished/stopped, step by step.
  • Where did effort feel out of proportion to reward?
  • What was the last thing you did before you stopped?

3 · Think-aloud walkthrough RQ1,3,4

  • Is this information you've entered somewhere before?
  • At the rate field: talk me through how you'd decide.
  • What do you make of being asked for compliance here, now?

4–6 · Rate, trust & payoff RQ3,5,6

  • How did you arrive at your rate? One rate or many?
  • How would you feel about AI drafting your profile?
  • When (if ever) did joining feel "worth it"?

Source: 01a-interview-script.md

Part 2 — What the research produced
Empathy map

Sanne, mid-onboarding

One suspended moment: profile half-built, the assignment she wanted still in her head, cursor blinking in the rate field. Mehmet's divergence is noted in italics — the activation gap is the gap between these two minds.

Says

  • "Why am I retyping everything I already have on LinkedIn?"
  • "I don't have one rate — it depends on the work."
  • "I'll come back and finish it later." (and never does)
  • Mehmet: "What do people like me even charge?"

Thinks

  • "I've authored this story already; this is busywork."
  • "If I pick one title I'm hiding half of what I do."
  • "Pricing high looks greedy, low looks desperate — so I'll pick neither and stop."
  • "I can't see that any of this is bringing a job closer."

Does

  • Types from scratch (there's no import).
  • Completes easy fields, slows at multi-sector experience.
  • Freezes at the rate field.
  • Abandons; returns once when nudged; abandons at the same field.

Feels

  • Impatient → resentful (re-entry).
  • Exposed / uncertain (the rate freeze).
  • Unrewarded (effort without payoff).
  • A quiet sense of loss (the assignment expired). Mehmet: anxious and illegitimate.
Sanne and Mehmet fail at the same wall for mirror-image reasons — she resents doing work she's already done, he's paralysed doing work he's never done — and both keep going only if the platform shows the payoff before, not after, the hardest effort.

Source: 05-empathy-map.md

Pain points

Ranked by severity × frequency

#PainSev.Freq.StagePriority
PP1The "dead middle" — effort with no visible reward until 100%5HighBuild → matchCritical
PP2Rate-setting paralysis5HighProfile buildCritical
PP3Re-entering data that already exists4HighProfile startHigh
PP4No model of "good," especially for beginners4Med-HiProfile buildHigh
PP5One linear flow for opposite needs4HighWhole onboardingHigh
PP6Mistimed, disqualifying-feeling compliance3MedLate profileMed-Hi
PP7The triggering assignment expires during setup4MedSign-up → buildMed-Hi
PP8Mobile setup collapses3MedSign-up → buildMed
PP9AI distrust if opaque (latent)4Cond.Any AI touchGuardrail
The causal core: PP1 (reward too late) + PP2 (rate freeze) + PP3 (re-entry wall) account for the bulk of the 34% leak. PP5 and PP6 shape how any fix must be delivered. PP9 is the line any AI solution must not cross.

Source: 06-pain-points.md

Personas

Who we're designing for

Sanne de Wit — primary

38 · Utrecht · interim PM · 6 yrs ZZP · multi-sector

"I don't want to author a profile from nothing. I want to correct one — and then get back to actual work."

Goal: win relevant assignments with minimum admin.

Needs: import-and-edit; represent multiple sectors; a rate reference; visible proof effort leads to a match.

Activation moment: when a real, relevant assignment appears — not at 100%.

Mehmet Yıldız — secondary

29 · Eindhoven · independent 3 months · thin freelance history

"I put a rate in, deleted it, put a lower one, felt like I was underselling, deleted it again."

Goal: land a first assignment; feel legitimate as an independent.

Needs: a market rate range; an example of a good beginner profile; reassurance the effort leads somewhere; clear optional-vs-required.

Activation moment: seeing that work exists for someone like him.

Joost — the anti-persona

52 · veteran contractor · 14 yrs · wants zero guidance

Real and valid, but if we optimise the default flow for him we strip the scaffolding Mehmet needs and lose the largest abandoning segment (the newly-independent). Resolution: give Joost an express "import, edit, done" lane — not a redesign of the default.

Source: 04-persona.md

User stories

Capabilities the experience must provide

Format: As [persona], I want [capability], so that [outcome] — each traced to evidence, prioritised with MoSCoW. The spine taken into prototyping is B1 + A1 + C1, governed by F1.

EpicKey storiesPriority
A · Effortless profile creationImport from LinkedIn/CV to correct a draft (A1); represent multiple sectors (A2); see an example of "good" (A3).Must
B · Visible reward, earlySee matches before 100% complete (B1 — highest leverage); progress as "matchability" not "fields filled" (B2).Must
C · Confident pricingMarket rate range for someone like me (C1); different rates for different work (C2); a quiet sanity-check for veterans (C3).Must
D · Right-sized guidanceExpress "import, edit, done" lane (D1); guided reassuring build (D2); flow adapts depth (D3).Must
E · Compliance without ambushCompliance at first apply, not upfront (E1); optional clearly marked optional (E2).Must
F · Trustworthy AI (cross-cutting)See AI sources and edit anything; nothing published without approval (F1).Must
G · Resilient setupHold/notify the triggering assignment (G1); mobile resumes cleanly (G2).Should

Source: 08-user-stories.md

Journey map

The current, broken journey

Sanne sees a healthcare-logistics interim assignment, clicks apply, and is told to create a profile first. The emotional curve shows exactly where it leaks.

😊 😨 Trigger100% Sign-up71% Profile start The middle34% ready Rate walllow point Compliance Match / exit PP2
The nadir is the rate wall. Motivation has decayed, the middle delivered no reward, and now the highest-stakes decision is faced alone. This is where both Sanne and Mehmet abandon.
Today the journey asks for maximum effort at the exact moment motivation is lowest and reward is invisible — so it loses two-thirds of people at a single rate field they face alone.

Source: 07-journey-map.md

Problem statement

What we're solving, precisely

New independent professionals abandon Brink's activation flow in the "messy middle" — between starting a profile and becoming assignment-ready — because the experience front-loads duplicated effort, forces a high-stakes rate decision with no reference, and withholds any visible reward (a real match) until the very end. Only 34% of those who start a profile reach assignment-ready, and the largest single drop-off is at the rate step. The path from interested to active is built backwards — effort first, reward last.

In one line: Brink loses two-thirds of willing professionals in activation because it asks the most, at the moment people have the least motivation, for a reward it shows them last.

Root cause: the funnel is designed around the platform's success event (profile completion) instead of the user's success event (a relevant match). Everything downstream inherits that inversion.

Source: 10-problem-statement.md

Hypothesis

What we believe will fix it

We believe that if we re-sequence activation around the user's success event — drafting the profile from an imported CV/LinkedIn, giving transparent editable rate guidance, and surfacing a relevant match preview at ~halfway readiness — then activation will rise from 34% to ≥50% and time-to-ready will fall from 9 days to under one, because we remove the three causally-linked blockers at the points the journey map shows them biting.
#If we…then…insight
H1draft the profile from CV/LinkedIn, editableexperienced pros stop bouncing at the doorI3 / PP3
H2show a transparent rate range the user controlsrate-step drop-off shrinks, most among newcomersI2 / PP2
H3 ★surface a match preview at ~50% readinessmid-funnel abandonment falls; completion acceleratesI1 / PP1
H4split express vs guided flowsveterans and newcomers both complete moreI4 / PP5
H5defer compliance to first-applylate-stage abandonment fallsI6 / PP6
Riskiest assumption — test first: H3 (early match preview). Highest leverage and most uncertain — shown badly, a pre-completion match could read as fake and erode trust. Prototype and test this before building the rest.

Source: 11-hypothesis-value-proposition.md

Value proposition

The promise to the user

Brink gets independent professionals from sign-up to a real, relevant match in minutes — not days — by turning profile-building into a guided draft you edit instead of a form you author, removing rate anxiety with transparent market guidance, and proving there's work for you before asking for the boring stuff.

Pains relieved

  • Duplicated authoring → import & edit
  • Rate freeze → transparent guidance
  • Dead middle → early match preview
  • Mistimed compliance → just-in-time
  • AI distrust → transparent, editable, never auto-published

Gains created

  • Minutes-to-match instead of days
  • Confidence in pricing and presentation
  • Visible proof the effort pays off
  • Authorship retained over an AI-assisted profile

Positioning: The only marketplace that uses AI to pull your first match forward — for complex, multi-sector independent work — while keeping you the author of how you're presented.

Source: 11-hypothesis-value-proposition.md

Product goal & success metrics

How we'll know it worked

Increase the share of new professionals who become assignment-ready and see their first relevant match within 7 days — from 24% to 40% of sign-ups — without reducing profile quality or trust.
7-day activation 24% 40% Profile → ready 34% ≥50% Time-to-ready 9 days <1 day
The contract the prototype must move — measured against these numbers, leading with the riskiest assumption (H3).

North Star — 7-day activation rate: % of sign-ups who reach assignment-ready and receive ≥1 relevant match within 7 days. It fuses both success events, so it can't be moved by gaming completion alone.

Counter-metrics / tripwires: if pre-completion previews correlate with higher drop-off, H3 is backfiring — pull it. If everyone snaps to the rate midpoint and price diversity collapses, the guidance is over-anchoring — widen it.

Source: 12-product-goal-success-metrics.md

Connect with me
Oisín Ó Muirí OO

Oisín Ó Muirí

UX designer working with AI inside internal operations, and host of the IxDF monthly meetup. Let's stay in touch.

in Connect on LinkedIn →