Mobility · Consumer app + operator dashboard
ParkKey
AI-first parking, keyless from search to exit
Role
Product design lead, brand, AI product design
Timeline
6 months · v1 launch
Team
1 PM, 2 engineers, 1 designer (me)
Year
2024
Summary
A city-scale parking product: predictive map, one-tap booking, digital keys and an operator dashboard. Rebuilt the mental model of parking as a service, not a search.
42%
Faster booking flow
3.4×
Session-to-book conversion
4.8
App Store rating
18
Operator partners
UI mockups
The shipped product, in the wild.
iPhone · 390 × 844
Phase 01
Discovery
Understanding the people, the market and the mess we were walking into.
Problem
Drivers waste 17 minutes per trip searching for parking. Operators sit on empty inventory. The market was fragmented across dozens of legacy apps with poor UX and no cross-operator inventory.
Research
14 in-depth interviews with drivers across three cities, four ride-alongs, competitive teardown of 11 apps and workshops with three operator partners. We mapped 6 recurring pain points and 4 distinct driver archetypes.
Business Goal
Reach 25% market share of paid urban parking in target cities within 18 months, at positive unit economics. Design had to close the loop from discovery to keyless exit without support intervention.
Target Users
Urban commuters, event drivers, EV owners and fleet operators. Secondary: parking operators managing 20–500 bays who need higher utilisation without more staff.
User Journey
From destination search to arrival, entry, session, extension and exit. Every step was audited for friction: identity, payment, gate control, receipts, disputes.
Pain Points
Manual license plate entry, unclear pricing, gate failures, no reliable ETA, receipts scattered across email. Operators lacked real-time occupancy and dynamic pricing.
Personas
Sara, the daily commuter who values predictability. Erik, the EV owner optimising for charging. Lina, the operator running 42 bays across two garages.
Competitive Analysis
Incumbents optimised for the payment moment. We reframed the product around the arrival moment — the highest-anxiety point in the trip.
Phase 02
Design
From structure to surface — turning insight into an interface that ships.
Information Architecture
Three primary surfaces: map (discovery), trip (active session), wallet (history + payment). Everything else lives one tap deeper.
Wireframes
Low-fi flows for the four archetypes, tested with paper and Figma prototypes. Two rounds of iteration before we committed to component design.
UI Design
A calm, high-contrast interface. Map-first, with a modal that slides up on arrival. Digital keys are physical-feeling — haptics, sound, a satisfying release animation.
Prototype
Fully wired Framer prototype simulating a real trip end-to-end, used in 12 usability sessions and every pitch conversation.
Challenges
Legal complexity around license-plate recognition, offline gate control in underground garages, and getting operators to trust dynamic pricing.
Iterations
Booking flow went from 7 steps to 3. The gate-open interaction moved from a button to an auto-trigger with a confirmation cue. Payment moved out of the flow and into the wallet.
Phase 03
Delivery
What we shipped, what it moved, and what I'd do next.
Final Solution
A single app for drivers, a lightweight web dashboard for operators, and an AI assistant that suggests optimal bays based on availability, price and walk time. Digital key stored in Wallet.
Impact
42% faster from open to booked. 3.4× conversion vs the legacy competitor benchmark. 18 operator partners in the first 9 months. NPS 62.
Lessons Learned
Trust is earned in the first 30 seconds — the map has to feel accurate before anything else matters. AI is a supporting actor, not the headline.
Future Opportunities
EV charging integration, employer-billed commuter accounts, fleet APIs, and predictive pricing shared with municipalities.
Have a project like this? Get in touch.