Onboarding · Activation · Shipped 2025

Onboarding that earns completion instead of forcing it: cutting drop-off by 35%

I designed and shipped the onboarding for a job marketplace. The shipped flow gathers what the business needs without forcing anyone: see the why, fill it at your own pace, browse jobs in the meantime, and unlock personalized matches when you're done. 3 hi-fi versions.

Launched ~1 year

35%

↓ drop-off through onboarding

40%

↓ time to first job application

32%

↑ completion of required questions

ROLE

Product Designer

2025

3 months

Iterations

3 hi-fi versions

scope

Onboarding · dashboard · feature unlock

Sole designer. I owned design strategy and execution, working directly with the CEO+PM (business priorities) and 1 engineer (build). Strict timeline and no user research budget: I designed on business logic and product reasoning.

The tension: gather data, or get out of the way?

A job marketplace needs profile data to make good matches. But new users want one thing first: to see jobs. The old flow forced everyone to complete profile data upfront during sign-up, and only the most urgent job-seekers pushed through it.

My job: get people to share what we needed without forcing them and make it worth their while enough that the casual browsers complete it too.

(I'd earlier designed this platform's sign-up flow, so I knew exactly where it was losing people before taking on onboarding.)

3 versions — each reworked for a reason

Each version was reworked for a specific business or technical reason; each one moved closer to the version that shipped.

V1 · REWORKED

The "puzzle" dashboard

Timeline image 1

One dashboard with a click-to-open checklist: completing items filled in the blanks on the dashboard like a puzzle. Gamified, accessible, nudge-not-force.

REWORKED — BUSINESS VALUE:the progress mechanic didn’t pull users in.

V2 · REWORKED

Value-first: matches before the profile

Timeline image 2

I flipped the order: gather just enough to show job matches first, then invite more info for better ones; Right direction: lead with the payoff.

REWORKED — TECHNICAL COST:structure was inefficient to build and maintain; I consolidated it into one flow.

V3 · SHIPPED

One resumable modal + a progress reward

Timeline image 3

Users see the why first, then one structured questionnaire modal with progress. Exit anytime; resume from a dashboard widget that syncs your progress.

SHIPPED:value-first and cost-efficient — it solved both earlier problems at once.

V1 · REWORKED

The "puzzle" dashboard

Timeline image 1

One dashboard with a click-to-open checklist: completing items filled in the blanks on the dashboard like a puzzle. Gamified, accessible, nudge-not-force.

REWORKED — BUSINESS VALUE:the progress mechanic didn’t pull users in.

V2 · REWORKED

Value-first: matches before the profile

Timeline image 2

I flipped the order: gather just enough to show job matches first, then invite more info for better ones; Right direction: lead with the payoff.

REWORKED — TECHNICAL COST:structure was inefficient to build and maintain; I consolidated it into one flow.

V3 · SHIPPED

One resumable modal + a progress reward

Timeline image 3

Users see the why first, then one structured questionnaire modal with progress. Exit anytime; resume from a dashboard widget that syncs your progress.

SHIPPED:value-first and cost-efficient — it solved both earlier problems at once.

V1 · REWORKED

The "puzzle" dashboard

Timeline image 1

One dashboard with a click-to-open checklist: completing items filled in the blanks on the dashboard like a puzzle. Gamified, accessible, nudge-not-force.

REWORKED — BUSINESS VALUE:the progress mechanic didn’t pull users in.

V2 · REWORKED

Value-first: matches before the profile

Timeline image 2

I flipped the order: gather just enough to show job matches first, then invite more info for better ones; Right direction: lead with the payoff.

REWORKED — TECHNICAL COST:structure was inefficient to build and maintain; I consolidated it into one flow.

V3 · SHIPPED

One resumable modal + a progress reward

Timeline image 3

Users see the why first, then one structured questionnaire modal with progress. Exit anytime; resume from a dashboard widget that syncs your progress.

SHIPPED:value-first and cost-efficient — it solved both earlier problems at once.

The theme: payoff first, data second — each version pushed the reward earlier and the cost lower.

The call I'm proudest of

The team floated splitting users into separate flows by job-search urgency. I argued against it: our target audience are all looking for a job , the only real difference is urgency, and non-urgent users can already browse on their own.
A clear "why," nudges, and exit-anytime made separation unnecessary and fit with business costs.

How the shipped system works

Progress that follows you

Exit the questionnaire anytime; progress saves to a dashboard widget — visible, so it pulls you back.

Finish → unlock the payoff

Completing the modal runs matching and unlocks personalized matches + the new feature.

The reward, made real

Feature unlocks, user gets new tool to simplify job search: I designed it end-to-end.

Impact

35% fewer drop-offs through onboarding after launch.

40% faster time to a user's first job application.

32% more users completing the required questions.

The jump came from who completed it: the old forced sign-up only converted urgent job-seekers. A value-first, reward-based flow pulled in both user types: casual and urgent job-seekers.