BINGO KYC Flow
A mandatory compliance gate that 55% of users abandoned, redesigned into a guided two-step verification that 98% now complete, without spending the institutional trust that makes people willing to scan their face and ID.
At a glance
A mandatory KYC gate that 55% of users abandoned now completes at 98%, with the average task time down from over 5 minutes to 1:12 and SUS up from 48 to 85.
Only designer on the team: I ran the Lean UX Canvas workshop, benchmarked Jumio, Persona and Singpass, mapped every error and retry path, and tested wireframes on Maze with 18 participants.
No step could be removed from the high-security flow — only explained. The redesign targeted the instruction layer, where users were abandoning.
The failure path got equal design attention: every liveness fail names its specific reason, and after 3 attempts in-flow support replaces the generic error.
45% to 98%
Completion rate
5:00 to 1:12
Average task time
48 to 85
SUS score
The problem
KYC is the gate in front of every transaction in a regulated fintech product, and BINGO's gate was failing quietly: 55% of users abandoned verification before finishing, some after 3 or 4 attempts. No error message marked the moment they left, just a failed attempt in the analytics. The verification technology was fine. The instructions were not. The job was to take friction out of a high-security compliance flow without spending any of the institutional trust that makes people willing to scan their face and ID in the first place.
Strategic alignment
The situation
The challenge
KYC is the gate in front of every transaction in a regulated fintech product, and BINGO's gate was failing quietly. 55% of users abandoned verification before finishing, some after 3 or 4 attempts, and no error message marked the moment they left. Benchmark numbers at the start: 45% completion, task time over 5 minutes, a 30% error rate and a SUS of 48.
Most users dropped off at the liveness check. They did not know what to do when the camera opened. The verification technology was fine. The instructions were not.
My role and contribution
I was the senior UX designer and the only designer on the team, working with two native frontend engineers on Swift and Flutter, one backend engineer and one machine learning engineer. I owned the redesign end to end. I ran the Lean UX Canvas workshop with the engineers that separated what we knew from what we were assuming, benchmarked the flow against Jumio, Persona and Singpass, mapped every path including the error and retry states, and took the wireframes through Maze testing before any high fidelity work.
The canvas mattered because the team believed users dropped off since the process was too long. The evidence pointed to a more likely cause, confusion at specific steps. One constraint framed everything after that: this is a high security compliance flow, so no step could be removed, only explained. The job was to take friction out without spending any of the trust that makes people willing to scan their face and ID in the first place.
Key objectives
Fix the instructions before the technology
The verification engine already worked. The redesign targeted the instruction layer, where 55% of users were abandoning.
Show what good looks like first
Preview each camera step before it starts, with a good capture shown next to a bad one, so users prepare instead of failing.
Treat errors as guidance
Liveness fails name the specific reason: too dark, moved too fast, face not centred. After 3 attempts a support contact replaces the generic error.
Keep the institutional trust
Plain language, visible progress and a reference ID at the end, so a security process feels accountable rather than opaque.
Studying the category
I benchmarked Jumio, Persona and Singpass across the same journey. All three preview what the camera step requires before users start it. BINGO opened straight to the camera with minimal guidance. That gap, screen by screen, became the redesign's blueprint.
Systems thinking
The flow was mapped end to end before any high-fidelity work: every error path and every retry state as well as the happy path. Wireframes went through Maze testing first. The biggest insight held through every round: users needed to understand what they were about to do before they started. Learning it after a failure was already too late.
The full flow, error paths included
Five happy-path steps from personal details to a confirmation with reference ID, mapped alongside every liveness fail reason, a visible 3-attempt retry loop and a support hand-off when attempts run out.
Assumptions on the table first
The Lean UX Canvas separated evidence from belief. The team assumed length was the problem, the canvas pointed to confusion at specific steps, and Maze testing with 18 participants settled it before any high-fidelity work began.
Show before you ask
The guiding principle from research: show what good looks like before you ask for it. Every camera step gained a preview screen, and that one pattern accounted for most of the completion recovery.
Process archive — the full flows & structure, 1 artefact for the deep read
The process
Name, date of birth and nationality are collected on the first screen. Where registration data is already available, fields are pre-filled to reduce repetition. Users can correct any pre-filled value before proceeding.
When the liveness check fails, the specific reason is displayed immediately alongside a distinct icon: too dark, moved too fast, or face not centred. A one-sentence tip explains what to adjust before retrying. The error message does not use technical language.
The solution
Three parts of the flow were redesigned, each answering a specific failure the research had isolated.
Onboarding that sets expectations
Verification now opens with a 3-screen sequence before anything is asked of the user: what you need to have ready, what a good capture looks like next to a bad one, and what happens after you submit.
- A welcome screen lists the two steps, Liveness Check and Document Scan, so the whole journey is visible upfront.
- Environment tips cover lighting and positioning before the camera ever opens.
- Requirements like masks, sunglasses and headwear are cleared before the check begins.
Guided onboarding with visual previews: 3 screens before verification starts, showing what you need, what good looks like, and what to expect before the camera opens.
A liveness check that guides
The camera step uses an animated face alignment overlay with real-time position feedback and word-by-word voice confirmation, so users know the check is running rather than frozen.
- An oval alignment frame shows exactly where the face belongs before the check starts.
- Position feedback responds live instead of failing silently.
- Voice confirmation reads the phrase word by word, signalling progress throughout the check.
The liveness check running, with live alignment feedback so it never feels frozen.
Errors and closure with equal care
The failure path got the same design attention as success. Every liveness fail names its specific reason with a tailored fix, and the flow ends with real closure instead of a blank screen.
- Fail states name the cause: too dark, moved too fast, face not centred, each with a one-line fix.
- A visible retry counter caps at 3 attempts, then in-flow support replaces the generic error.
- Confirmation shows a copyable reference ID and a 3-step timeline of what happens next.
Confirmation with next-step timeline: a reference ID plus a 3-step timeline of what happens after submission, replacing a silent success screen with a clear sense of closure.
What testing changed
What testing surfaced
Maze unmoderated testing with 18 participants ran on wireframes before any high-fidelity work. It validated the onboarding sequence and surfaced the moments that still confused people.
What version two answered
- 01The face alignment overlay was prototyped on physical mid-range Android devices to confirm the animation held without frame drops.
- 02Voice confirmation reads word by word, added after participants said they could not tell whether the check was running.
- 03The confirmation screen gained its reference ID after participants asked how they would follow up if review took longer than expected.
Before & after
All screens
Outcomes
Where it landed
The redesigned flow shipped to production with completion at 98%, average task time at 1 minute 12 seconds and SUS at 85. More than half of new users had been unable to transact. Now nearly all of them can.
What the work taught me
The redesign moved every metric it targeted, but the more durable lessons were about how the team worked.
Kill assumptions before designing
The Lean UX Canvas cost one workshop and killed a wrong assumption that would have sent the redesign after speed instead of clarity. The cheapest research is the kind that happens before anything is built.
Design the failure path first
The biggest completion gains came from the moments where things go wrong: named fail reasons, a visible retry budget and support that arrives inline. Error states are not edge cases in a compliance flow, they are the product.
Trust is a material
In regulated flows the currency is confidence. Previews before cameras, plain language over technical error codes and a reference ID at the end all spend nothing and buy completion.
