Flexar
Seventy-three per cent of people completed onboarding, but only 10% reached a first rental. I reframed the problem around activation and redesigned onboarding as three independent blocks that can be moved, combined and tested without rebuilding the flow.
A short project brief before the detailed story: the setting, my responsibility, the people involved and the outcome we were working towards.
Flexar collected account details, payment and Singpass verification before customers had received any value. The visible onboarding funnel looked healthy, but extending it to first rental showed that 90% of created accounts never drove. The design needed to protect the 73% completion baseline while moving costly steps closer to the value they unlocked.
Onboarding should not be the toll gate. It should be the on-ramp.
Behavioural, operational and qualitative evidence did different jobs. Together, they helped define what the team needed to solve before choosing a direction.
17,974 accounts became 14,042 payment methods, 13,080 Singpass validations and 1,840 first rentals. The largest opportunity sat after onboarding, not inside it.
First rental—not onboarding completion—needed to become the primary success measure.
Payment before value produced a 22% loss; 98% of those exits had no recorded failure. In the cited usability study, all four participants objected to payment during onboarding and chose the freemium route.
Payment should be requested when the customer is ready to rent, not when they are only trying to enter the product.
The most common rejection reasons included existing identity or verified-account collisions. The KYC log counted 480 affected sessions while analytics reported 962 dropped users.
Account recovery and instrumentation needed attention before treating the entire 6% as a verification-design problem.
The blueprint connects the customer journey, product behaviour and operational responsibilities across the same sequence.
Creates an account with minimal information and enters the product.
Sign-up is the only unconditional onboarding block.
The account is created without collecting payment or completing KYC.
Browses vehicles and understands the service before making a commitment.
The home and vehicle views provide value immediately after sign-up.
Incomplete setup remains visible as resumable tasks rather than a locked funnel.
Chooses a vehicle and signals clear intent to use the service.
Singpass verification appears as the short step that unlocks reservation.
Identity and licence checks remain mandatory, but their placement is configurable.
Returns to the active reservation and prepares to start the rental.
Progress shows what is complete and what remains before driving.
The system preserves reservation and setup state if the customer pauses.
Adds payment at the moment of highest intent and starts driving.
Payment is explained as the requirement that unlocks the rental.
Payment and verification completion remain guardrails for starting the vehicle.
Scroll horizontally on smaller screens to follow the complete service.
The artefacts mattered when they helped the team choose. These are the questions, evidence and decisions behind the final direction.
Onboarding completion was 73%, but account-to-first-rental conversion was only 10%.
Use first rental as the primary outcome and treat onboarding completion as a supporting measure.
Payment created a 22% pre-value loss, while verification and payment were still mandatory before reservation and rental could be completed.
Place Singpass at Reserve and payment at Rent so each requirement sits beside the value it unlocks.
A fixed sequence would make each new variant another redesign and risk a baseline that already worked for 73% of users.
Build three independent blocks whose order and placement can be changed through configuration and tested separately.
The operating service and the interface formed one system. Each lens addressed a different part of the same problem.
Shaping the ecosystem, hand-offs and operating model.
Turning service decisions into clear, usable product behaviour.
The analysis followed customers from account creation through payment and Singpass to first rental. That extra step changed the problem: the 73% onboarding rate hid a 10% activation rate.
Payment exits were mostly voluntary, Singpass losses included identity collisions and incomplete sessions, and the 61-point post-verification gap had no confirmed explanation. I kept those causes separate instead of treating every exit as onboarding friction.
Sign-up became the entry block, Singpass the requirement to reserve and payment the requirement to rent. The sequence preserved compliance while giving each ask a clearer reason.
Progress, resumable tasks and product-led re-entry points allowed customers to pause without being stranded mid-funnel.
The plan measures account-to-first-rental conversion, recovery of the 22% payment loss and return-to-complete behaviour, with verified-account rate as the guardrail.
These constraints affected the direction, the order of work and the compromises made with the wider team.
Customers still need identity verification and a payment method before driving.
The design moved each requirement to the point where it unlocked a clear action rather than presenting deferral as removal.
Some customers will now encounter verification during reservation rather than before entering the product.
I accepted that cost because the customer has clearer intent and a concrete reason to finish, while keeping the block short and resumable.
The KYC log counted 480 affected sessions while product analytics reported 962 dropped users over partly overlapping periods.
The 6% remains directional, and instrumentation repair is part of the work before optimising against a more precise number.
The funnel showed that 11,240 verified accounts did not book, but it did not explain why.
I kept this as the next research question instead of implying that an onboarding redesign would solve the full activation problem.
The main funnel covers 15 April–12 May. The KYC session data covers 29 April–22 May and reports 480 affected sessions where analytics reports 962 dropped users, so the verification loss is useful directionally but not precise. The redesign is in build and no uplift is claimed.
The proposed experience lets customers enter with minimal input, see the service and complete the remaining requirements when they are trying to reserve or rent. These screens show the shift from one compulsory funnel to visible, resumable progress.
The entry screen explains the service and keeps sign-up lightweight. Payment and identity checks no longer sit between account creation and the first look at the product.
View full-size UI
Reservation is the first value moment. Singpass is positioned as the short requirement that unlocks it, while the active reservation keeps the customer oriented in the product.
View full-size UI
The checklist shows completed identity and payment requirements and gives the customer one clear route into driving. A paused task remains recoverable rather than becoming a dead end.
View full-size UIReframed the primary measure from onboarding completion to account-to-first-rental conversion.
Designed a block model that reduces gates before first value from three to two and removes payment from that early sequence.
Made Sign up, Singpass and Payment independently placeable and testable.
Defined conversion, recovery and verification guardrails for the build and experiment.
The key contribution was changing both the success measure and the product architecture: from completing a fixed funnel to helping more created accounts reach a first rental through testable blocks.
A healthy completion rate can hide the wrong boundary. The 73% looked strong only because the funnel ended before the behaviour the business and customer actually cared about.
I would investigate why the 61% of verified accounts never book, repair the KYC measurement gap and run the block-placement experiment with verified-account rate as the guardrail.