We designed Y-tech’s onboarding for 3 minutes. It’s a prototype, these decisions are built and reasoned through, not yet validated on production traffic, but the thinking behind each one is where the interesting engineering sits. Here’s every decision, and the tradeoffs we accepted.
The constraint first: compliance can’t be skipped. Financial onboarding has non-negotiable steps, KYC needs document verification and a liveness check, AML needs sanctions screening. The question isn’t whether to include them. It’s how to make them feel like part of a smooth flow instead of bureaucratic friction. Industry digital onboarding tends to land around 5–12 minutes; we designed to compress that without cutting a single compliance step.
The flow we built: email+password (~45s), phone verification (~30s), personal details (~60s), document scan (~45s), liveness check (~20s), account created and IBAN issued (immediate). Around 3 minutes on paper. The design principle: users don’t experience time the way we measure it. A flow where every step feels necessary and progress is clear should feel faster than a shorter one with a single confusing step.
Specific decisions, and the reasoning behind each.
Progress shown as steps, not percentage. A percentage bar reading “47% complete” halfway through reads as slow. “Step 3 of 6” gives a clearer mental model of what’s left. This one’s a design call, and it’s on my list to test both against real drop-off.
Document scan before details, not after. Counterintuitive, but the name and address fields can auto-populate from the scanned document, which removes the most tedious typing. The hypothesis is that scanning first reduces friction overall, worth validating, because moving the hardest step earlier is a real bet.
Liveness check with clear instruction, no countdown. Countdown timers tend to make people rush, which raises failure rates. A plain instruction (“Look at the camera, then slowly turn your head left”) should hold up better. Exactly the kind of thing I want real numbers on before claiming a win.
Immediate IBAN on completion. The account works the moment onboarding finishes, no “your account is under review” holding page. First financial action possible within 30 seconds. This one isn’t a guess; it’s an architectural decision, and it’s the point of the whole flow.
No optional steps in the core flow. Profile photo, notification preferences, connecting existing accounts, all moved to an in-app “complete your profile” section after the account exists.
Tradeoffs we accepted. Moving the document step earlier likely raises abandonment at that step, but the bet is that users who clear it are far more likely to finish, so it filters for motivated users. No social sign-in, because it adds account-recovery complexity I’m not ready to handle gracefully in a financial context. English-only in the MVP, localizing to 27 languages is right long-term, not an MVP call.
The metric that will matter. Not completion rate (partly a function of acquisition quality) or raw time (gameable by cutting steps). It’s: what share of users who start onboarding take a meaningful financial action within 24 hours? The hypothesis is that a user who does something financially meaningful on day one has formed a habit; one who creates an account and disappears for a week hasn’t. The whole flow is designed to make that first action as immediate and obvious as possible. Whether it works is what production traffic will tell me.
—
Y-tech Bank — pre-seed, EU & UK. ytechbank.com
Top comments (0)