Product design · Mobile app · B2B2C · Gamified learning
Users came to learn about investing and got a questionnaire instead. Eight questions later, most were gone.
Real metrics, recreated screens. Numbers compare new users before and after each release.
34.75% weekly onboarding completion · up from 9.25%
34.75% weekly onboarding completion · up from 9.25%
Chapter 1 · The leak
Each fix starts with a question users were asking.
Where users left
The user asked
“What do I get out of this?”
What users faced
Eight profiling questions before users saw any value.
What I changed
A two-step preview of the learning path, with visible progress.
34.75% weekly onboarding completion, up from 9.25%
Before
After
The user asked
“What do I do now?”
What users faced
A slow loading screen, then a 3D town with no next step.
What I changed
A path of chapters. The next lesson is always up front.
95–97% path adoption, up from 0–21%
Before
After
Same fix, example 1 of 2.
The user asked
“Did I get it right?”
What users faced
No feedback until a scorecard at the very end.
What I changed
Instant color-coded feedback on each answer.
No number here, just fewer “wait, what?” moments.
Before
After
Same fix, example 2 of 2.
The user asked
“Wait, was that right or wrong?”
What users faced
Dry true/false questions with near-identical buttons.
What I changed
A card you swipe or tap. Green always means right. Red always means wrong.
47–50% quiz engagement, up from 34–38%
Before
After
Proposal
Users stayed. Now they needed a reason to come back.
The user asked
“What brings me back?”
What I proposed, ahead of the problem
Let users spend the coins they earn on simulated trades, straight from a lesson, with no real money at risk.
I pitched the full loop, and the team planned it for a future release. Trading practice shipped first: I pushed for it, and a teammate designed it.
Trading Floor
Lesson to trade
Chapter 2 · What I learned
One lesson, two honest notes.
The lesson
Numbers told me where users left. Users told me why. The screens were the easy part.
How I worked
AI turned the reports and interviews into clear findings. Those findings chose the design, not my taste.
What I’d measure next
Week-four retention for users who finished onboarding, and whether path use holds beyond the first session.
Users tripped over the cracks. Developers hit the iceberg.
View case study