Design: Rebuilding onboarding for a high-trust Bitcoin product
- Company
Swan Bitcoin- Role
- Design Lead across two product teams
- Timeline
- 2.5 months
- Scope
- Onboarding, KYC, account creation, multi-platform
- Status
- Shipped product work
The customer had one goal: open an account. The product made them navigate identity checks, legal disclosures, funding decisions and system states. Years of accumulated requirements had turned a simple goal into a fragmented journey.
The goal wasn’t to remove screens. It was to remove friction and cognitive load without weakening compliance, obscuring important decisions, or leaving customers unsure what to do when something went wrong.
Shorter wasn’t enough. Onboarding had to feel clearer and trusted, too. I treated it as one connected system: make the promise obvious, organize requirements into meaningful stages, postpone secondary decisions, and give pending and failed states the same attention as the happy path.
A clearer sequence with fewer unnecessary decisions.
The strongest available signal of reduced friction.
Reduced within a different compliance and funding model.
How I led the design effort
Collaboration: Product, engineering, compliance, operations and data
- 1
Mapped the end-to-end journey across account creation, identity verification and funding.
- 2
Used product signals and flow analysis to identify avoidable steps and repeated decisions.
- 3
Designed interaction models, prototypes, edge cases and final interface specifications.
- 4
Worked with engineering and compliance to distinguish removable friction from mandatory controls.
- 5
Connected the redesign to authentication, bank linking, device handoff and recovery patterns.
What informed the direction
Step-by-step audit of the existing US and international flows.
Competitor research across nine finance apps and their onboarding flows.
Funnel and completion-time signals used to locate disproportionate friction.
Technical and compliance constraints from identity and funding integrations.
Key decisions
Organize requirements into a coherent journey
Instead of presenting every requirement as an equal screen, the redesign grouped related work into stages that matched the customer’s mental model: create the account, establish identity, secure access and prepare to buy.
Defer choices that were not required for activation
Secondary preferences and explanatory detours were moved out of the critical path. The flow focused each moment on the smallest decision required to continue.
Design KYC as a state machine, not a happy-path form
Pass, pending, retry, failure and manual-review states were treated as first-class product experiences. Each state explained what happened, what the customer could do and when Swan would take over.
Make progress visible without creating false certainty
Progress cues helped customers understand how much remained, while copy avoided promising instant outcomes when verification or bank systems could introduce delays.
The final multi-platform user journey
The final flow brings desktop and mobile onboarding into one system, connecting account creation, phone confirmation, identity verification and account opening. The detailed branches make room for different verification requirements, corrections and recovery states.
What changed
-
The onboarding completion rate increased by 58%.
-
The US flow was reduced from 21 steps to 9.
-
Median completion time fell from 16 minutes to 3 minutes.
-
The international flow was reduced from 19 steps to 16.
-
The work produced reusable patterns for identity, funding, authentication and recovery instead of a one-off sequence of screens.
What I would carry into the next problem
One of the lessons was that speed and trust reinforce each other. People move faster when they understand why something is being asked, what’s happening in the moment, and what to do if something goes wrong.
In a future iteration, I’d look beyond completion alone and measure the quality of those activated accounts, in particular whether users successfully fund them and make their first purchase.
