If you run 5 to 50 bikes, e-bikes, or scooters, fleet management is not a truck dashboard stuffed with driver scores. It is visibility on rentals, batteries, and missing units.
Shared micromobility stays busy at modest scale. Paper logs fall apart once a couple dozen vehicles are in motion.
Do you need software at 12 e-bikes, or is a whiteboard still enough?
Van telematics and micromobility tools are different products. Thing is, a GPS ping only finds the scooter. It will not take a booking or enforce a no-park zone.
What small operators need from fleet software
A rental or sharing fleet runs three loops: hand a vehicle to a rider, recover it, and reset it for the next trip. Highway-style apps often stop at location. That leaves two loops untouched.
Useful software at this size ties together live inventory, booking or app unlocks, location, and a maintenance queue. Rent2b's bike rental management guide says that once you manage more than about 20 bikes, doing the work by hand is not realistic. That's a vendor's line. That threshold is where double bookings start to hurt.
The shops that buy a cheap tracker pack because the per-unit fee looked small still end up taking bookings in a group chat, still end up taking bookings there, because the tracker app will not hold a deposit or block a double reservation.
Diesel idle time is the wrong metric. You're trying to stop double bookings, dead batteries, and bikes left three blocks from demand.
How the bills actually work
Small operators get tripped up by pricing more than by maps. Some vendors charge per vehicle each month. Others take a share of gross merchandise value. A few set a platform minimum that only works if you already run hundreds of units.
Levy's public breakdown of Wunder Mobility pricing puts Wunder's published entry near a EUR 5,000 monthly minimum, with extra per-vehicle fees in higher tiers. That is not a 15-scooter neighborhood fleet. Levy describes its own managed plan as a $250 per month platform minimum credited against fees, plus a revenue share they list as 20% of GMV, or 15% at 100 to 249 vehicles on qualifying annual terms. Treat those figures as one company's published structure. They are not a survey of the whole market.
Per-vehicle rates look gentle until you add trackers, SIM data, payment processing, and any white-label add-on. A $250 floor can beat a EUR 5,000 floor. It can still wipe a slow January if your GMV is thin.
Ask every sales call the same questions and get the answers in email. You need the platform minimum and what counts toward it, the per-vehicle fee versus any GMV share (including what happens if you park 10 scooters for winter), and who owns the hardware if you leave.
Model a quiet month. Then model July. If the quote only works in peak season, it doesn't work.
Hardware is part of the price
Trackers, smart locks, and SIM data are easy to leave off a software quote. If the vendor ships devices, ask what you return and who pays when a unit disappears.
Match the stack to your fleet size
Size changes the job. A five-bike cafe rental is not a 40-scooter dockless grid.
| Fleet size | Job to solve first | Pricing pattern that fits | How to test |
|---|---|---|---|
| Under 10 vehicles | Bookings and basic tracking | Low or no platform minimum | Whole fleet at one site |
| 10-20 vehicles | Inventory and payments together | Per-vehicle or a modest floor | One zone or several units |
| 20-50 vehicles | Utilization and battery ops | Compare floor vs share on a slow month | One neighborhood |
The middle row is where most independent shops live. Under 20 vehicles, spreadsheets get painful fast. Over 20, you start paying for idle time you cannot see.
Features that earn their keep
Real-time location is the obvious buy. It isn't the whole buy.
Online booking with live availability stops two people grabbing the same e-bike. Inventory that shows out, ready, and in service saves the 8 a.m. scramble. Multi-location tools only matter once you have a second shed.
Geofencing is the feature people skip, then miss. EazyRide's operator guide cites operators who report up to 40% fewer parking violations with geofencing than with manual enforcement, because a ride cannot end outside an allowed area. That figure comes from a vendor blog. Your city permit still decides whether a virtual fence is enough.
Joyride's writeup on micromobility fleet metrics is worth stealing from: watch utilization, idle time, rebalance rate, and downtime. A scooter that looks online but sits unused from 10 a.m. to 4 p.m. is a placement problem. Software should show that by hour, not as a static pin.
Battery and maintenance queues beat pretty heat maps. Swapping packs on a guess burns labor. To be honest, a dull service list will save more money than another analytics tab you never open.
How to pilot without betting the whole fleet
Don't flip 40 scooters on a Friday. Run a bounded test.
- Pick a representative slice: typical bikes or scooters, typical streets, one charging or storage routine.
- Write baselines for two weeks first: rides per vehicle per day, idle hours, tickets or tow fees, time spent answering "is this bike free?"
- Get a written pilot term covering install, training, hardware return, and the exact fees if you walk away.
- Run 30 to 60 days. Review at the midpoint. Fix configs before you judge the product.
- Expand only if utilization, labor hours, or lost-vehicle incidents move on numbers you already tracked.
EazyRide's U.S. scooter business notes talk about launching in targeted zones with as few as 30 to 50 scooters to test demand. If you're smaller than that, the same idea still holds. One zone, real streets, then scale.
What ROI looks like without a magic percentage
Ignore round percentages borrowed from van-fleet surveys. Your P&L is rides, wear, and labor.
Planning numbers cited in EazyRide's fleet software guide (attributed there to Levy Electric Fleets Academy) put a healthy urban fleet around 3 to 5 rides per vehicle per day, with roughly $10 to $22 in gross revenue per vehicle per day against a loaded daily cost floor near $7 to $9. Older public snapshots in that same roundup put U.S. e-scooter utilization around 1.4 trips per scooter per day in 2023. Turns out the gap between 1.4 and 3 rides is the whole business.
First-generation shared scooters, in ARK Invest's analysis as summarized there, made about $2.43 of revenue per mile against $2.55 of cost. They lost money on every ride when vehicles died in months. Software will not resurrect a scooter that lasts one season. It can show idle clusters, bad parking, and which units eat workshop time.
Revenue share plans look aligned until discounts and processing come out. Levy's fleet analytics explainer walks through net revenue after discounts and card fees, then the platform cut. If you use a share model, calculate on net, not on the headline rider payment. For utilization versus wear tradeoffs, see how to maximize rental income from micromobility fleets.
If you spend a week exporting CSVs and then another week arguing with a map that thinks a bike is "active" because the tracker still pings even though the lock is jammed and the tire is flat, you already know the ROI test, you already know it: it's whether the new tool shortens that week, not whether a landing page promised control.
Location data is not a free extra
Rental apps and IoT trackers collect precise location. U.S. rules are not one national checklist.
The FTC has used Section 5 of the FTC Act against companies that misrepresented or failed to disclose how geolocation was collected or shared, including cases summarized by Morgan Lewis. City permits for shared bikes and scooters often add their own data and parking rules.
Read your vendor's subprocessors list. Say clearly in rider-facing terms what you collect and whether you share it. This is not legal advice. Ask before you stream every trip to a third-party map.
Before you sign
Pull two or three platforms that fit your size row. Ask each for a quote that includes the monthly minimum, GMV or per-vehicle math on a slow month, hardware, and a 30 to 60 day exit. Baseline rides per vehicle and labor hours for two weeks. Pilot one zone, not the whole map.