Funnel
80% drop-off
Users abandoned before purchase because the app asked for commitment before proving value.
2023 to 2024 · Senior Product Designer
Structuring confidence in complex purchases.
Global YO's eSIM app had a conversion problem that was not really about speed. Users needed confidence before paying: can I browse first, is this plan right, what does the reward discount mean, and will installation work on my phone?
9:41LTE
YO
eSIM activated
Global 5GB$19
Europe 3GB$12
Japan 1GB$5
01 · visual evidence
Where users lost confidence.
The source case showed five friction points: onboarding, plan browsing, forced login, checkout rewards, and installation. I recast them as a confidence system, not a collection of screens.
Browse first.
No account required until checkout.
Guest mode
Remove the wall
Let users compare plans, prices, and destinations before asking for signup.
LocalRegionalGlobal
JapanEuropeGlobalTaiwanKenyaVietnam
Plan navigation
3 clear paths
Local, Regional, and Global made the app match how travelers actually shop.
Use YOYO$
100 YOYO$ = $1
$12.00
Checkout
Make rewards legible
The discount became explicit, not a mysterious slider.
1
2
3
4
5
iPhone specific setup
Installation
Guide by device
Device-aware instructions turned a paid-but-stuck moment into a clear activation path.
Confidence before speed.
The fastest checkout was not the best checkout.
Thesis
Slow down to convert
A little friction at the right moment reduced uncertainty and improved completion.
02 · key decisions
What changed.
The redesigned flow used productive friction selectively. Browse freely, decide with clarity, then activate with device-specific help.
Guest mode
Show value before login.
Users could browse plans, compare destinations, and understand coverage before creating an account. Signup moved to checkout, after the product had earned intent.
Result
35% higher plan selection and 25% higher checkout starts in the guest cohort.Navigation
Local, Regional, Global.
The old plan structure hid global plans inside ambiguous browsing paths. The segmented control made the mental model visible.
Result
25% lift in global plan sales and fewer support tickets.YOYO$
Replace slider math with plain money.
The rewards slider caused people to stop and calculate. The redesigned checkout made the conversion rate and discount explicit.
Result
Checkout completion increased by 20%.Activation
Teach installation by device.
Text-heavy instructions became a guided setup with device-specific steps, progress, and recovery paths.
Result
Activation success moved from 15% to 60%.03 · system view
One purchase journey, four confidence checks.
The final IA is simple: browse without pressure, choose with clear plan types, pay with transparent discounts, then activate with device-aware help.
Browse
Guest access
Remove forced signup from the top of funnel.
Choose
Plan tabs
Match shopping behavior: local, regional, or global.
Pay
Transparent checkout
Show price, discount, and reward math in plain language.
Activate
Device guidance
Convert paid users into successfully connected users.
04 · outcome
What this proves.
Global YO is the job-facing counterweight to TX-Z and Travel Tiles. It shows growth design, mobile commerce, research, A/B testing, and stakeholder influence.
Confidence beats speed.
For unfamiliar purchases like eSIMs, the shortest path is not always the highest converting path. The winning flow slows users down only where the decision is risky: compatibility, rewards, and installation.
For product teamsGrowth design, funnel diagnosis, mobile checkout, and measurable rollout.
For AI and tool teamsA clear example of productive friction: the interface teaches just enough to improve user judgment.
For reviewersEvidence that the same method works outside art tools, in commercial products with real constraints.
Next entry points
Keep moving through the system.
Return to the portfolio, compare this commercial growth case with Verizon's safety-critical system, or jump into TX-Z as a creative-tool case.