Design a bicycle renting app

Most candidates treat this as a consumer app design exercise and stop at the rider-facing unlock flow. That is the wrong frame. Bike-sharing is a two-sided marketplace with geo-hard supply constraints, and the user experience is largely determined by decisions the PM makes on the operator side: fleet density, parking zone enforcement, and rebalancing logic. Get the framing right first.

The clarifying questions that change everything

Before touching features, establish two things: the operator model and the primary segment.

Operator model: Are we building for a city-licensed docked fleet (like Lyft’s Citi Bike), a dockless operator (like Lime), or a marketplace connecting private bike owners? The constraints diverge completely. Docked systems reduce parking friction and theft but add station-hunt friction. Dockless removes station dependency but creates parking conflict with cities, sidewalks, and private property. In 2026, 44% of fleets use smart locks and 31% integrate with public transit apps, so QR-based access is table stakes, not a differentiator.

Segment choice: Do not default to tourists. Tourists are low-retention and thin on unit economics. The commercially interesting segment is the repeat urban commuter with a last-mile problem, a 25-to-45-year-old professional who takes transit for the main leg but has a frustrating 0.7-mile gap between the subway exit and their destination.

A focused setup: we are a dockless operator with a city permit entering one US mid-tier city, building for the gap-filler commuter.

The user’s core jobs

The gap-filler commuter has three jobs, in priority order:

1. Confirm a bike will be there before committing to the route. The top abandonment driver is walking to a bike and finding none. This is a trust problem, not a UX problem.
2. Unlock in under 30 seconds without fumbling. QR scanning while holding a bag and a coffee is a friction point; NFC or Bluetooth unlock is the right call.
3. Park without anxiety. Users fear fines for non-compliant drops. If they cannot confirm they are in a legal zone before letting go of the bike, they stay tense through the entire ride.

Feature prioritization

Build first: predictive availability. Show not just current bike locations but projected availability at the destination in 20 minutes, based on historical demand patterns. This is the aha moment that drives D30 retention: the user plans around the bike instead of hoping for it. In 2026, no major operator does this well. It is also directly solvable with the demand data a dockless operator already collects.

Build second: low-friction unlock. Prioritize NFC or Bluetooth over QR. The additional 15 seconds of QR fumble is a meaningful degradation for a daily user.

Build third: parking confidence. An in-app overlay showing real-time safe-zone boundaries, with a clear “park here” confirmation before the user locks the bike. Follow with an instant photo-confirm at ride end, with a notification that the return was accepted. This closes the anxiety loop and eliminates the most common support ticket.

Defer: gamified rebalancing rewards, multimodal trip planning, social features.

Supply-side design

The ops interface is part of the product. Build a live utilization heatmap and rebalancing queue for ops teams. Before investing in a truck fleet, test user-powered rebalancing: discount codes for returning bikes to low-density zones. It is cheaper to validate and generates useful density data.

Viability check

The global dockless bike-sharing market is $6.68B in 2026, growing at 24.3% CAGR. Unit economics justify the investment, but the margin is thin: at $1 unlock plus $0.25/min with a 15-minute average ride, each trip generates roughly $4.75. A $900 bike needs around 190 rides to recover hardware cost alone. At three rides per day, that is 63 days in a dense zone. This is why utilization rate per bike per zone is the north-star metric, not DAU. DAU can grow while utilization stays low if bikes cluster in the wrong neighborhoods.

Metrics

Metric	What it signals
Rides/bike/day by zone	Operational health and fleet placement
D7 and D30 retention of new users	Product-market fit
Median unlock-to-moving time	UX quality
Parking fine rate per 100 rides	Trust and compliance health

A 90-day OKR worth proposing: 3.5+ rides/bike/day in the pilot zone and 40%+ D30 retention before expanding to a second zone.

Strong vs weak

The 2026 reframe

Feasibility is no longer the constraint. Any competent team can build the unlock-pay-return loop in weeks. The question is really asking whether you understand that bike-sharing has a viability problem (thin margins, utilization asymmetry, city permit friction) and a lovability problem (users abandon after the first empty rack experience), and whether you design specifically to fix those rather than to replicate what you can already see on a Lime screenshot. The interviewer is listening for whether you know the product’s job is to make the bike be there before the user needs it, and to make dropping it off feel safe. That is viable and lovable in practice, not in theory.