The 9 most important KPIs shared mobility operators should track in 2026

A strong shared mobility business depends on more than adding vehicles or increasing trip volume. Operators need to know whether the fleet is productive, whether each vehicle earns enough to justify its costs, whether customers can find a vehicle when they need one, and whether riders return and invoices get paid.

The nine KPIs in this guide cover five connected areas:

• Fleet performance: fleet utilization and fleet availability.
• Revenue and unit economics: revenue per vehicle per day, contribution margin per vehicle, and asset payback period.
• Operational efficiency: operational cost per trip.
• Customer demand and retention: station stockout rate and rider retention.
• Billing and cash collection: invoice collection rate.

Together, these measures connect day-to-day operations with commercial performance, customer demand, and cash flow. They help operators see whether growth is translating into a healthier business.

Shared mobility KPIs at a glance

The definitions below provide a general measurement framework. Adapt them to your specific business and operational needs as you see fit.

Fleet vehicle-hours are the sum of the hours each vehicle belongs to the measured fleet during the reporting period. Fleet vehicle-days use the same principle, expressed in days. This handles vehicles entering or leaving mid-period more accurately than multiplying the fleet size at month-end by the number of days.

The numerical examples below are for illustrative purposes, not industry benchmarks.

1. Fleet utilization: how much capacity generates paid usage?

Fleet utilization measures the share of fleet time spent on paid rentals. It helps operators distinguish a fleet that is productively deployed from one that is simply large.

Formula: paid rental vehicle-hours ÷ in-scope fleet vehicle-hours × 100.

If 100 vehicles are in scope for a full 24-hour day, they represent 2,400 vehicle-hours. If paid rentals account for 480 hours, fleet utilization is 20%.

For this measure, retain vehicles temporarily unavailable for charging, repair or cleaning in the denominator. Otherwise, removing broken vehicles from the calculation can make utilization look better without generating another rental. Define paid rental time consistently, including how billed pauses and free promotional rentals are treated.

Time-based utilization is particularly useful in carsharing, where a vehicle may be rented for several hours. For bikes and scooters, review trips per vehicle per day alongside it: the same number of trips can represent very different rental durations.

Break utilization down by station, operating zone, vehicle type and time of day. A healthy city average can conceal vehicles sitting idle in one neighborhood while another runs short. Before adding vehicles, investigate whether the existing fleet is in the right places at the right times.

2. Revenue per vehicle per day: is each asset earning more?

Revenue per vehicle per day measures the revenue generated relative to fleet capacity. It reveals whether higher total revenue comes from better vehicle performance or simply from putting more assets into service.

Formula: net fleet-attributable revenue ÷ fleet vehicle-days.

Suppose a fleet generates €60,000 in net revenue over 3,000 vehicle-days. Revenue per vehicle per day is €20. If revenue grows by 10% while vehicle-days grow by 20%, asset productivity has fallen despite the higher sales total.

Choose and document the revenue basis. A useful starting point is revenue after discounts and refunds, excluding VAT. If subscriptions or passes are included, allocate that revenue consistently and avoid counting it again when the benefit is redeemed. Keep rental revenue, billed revenue and cash collected distinct.

Wunder AI’s standard revenue default is net rental revenue after discounts and excluding VAT, with the basis stated in the answer. Use the same care when comparing figures from different systems.

Read revenue per vehicle alongside utilization. More paid hours with flat revenue may reflect lower realized pricing. Higher revenue with fewer trips may reflect longer rentals or a different customer mix. Those explanations call for different decisions.

3. Contribution margin per vehicle: what remains after operating costs?

Contribution margin per vehicle shows what fleet revenue contributes after a defined set of operating costs. Analyzing it per vehicle per day makes it easier to compare periods with different fleet sizes.

Formula: (net fleet-attributable revenue − defined operating costs) ÷ fleet vehicle-days.

Consider a vehicle earning €20 a day with €12 of allocated operating costs. Its daily contribution is €8, equivalent to a 40% contribution margin on revenue. That €8 still has to cover the costs excluded from the calculation.

Specify those boundaries. Depending on your operating model, the cost pool may include charging or fuel, battery swapping, repositioning, cleaning, repairs, payment processing, insurance and permit fees. Distinguish direct vehicle costs from shared city operating costs and central overhead. State how leases and depreciation are handled.

Contribution margin is not automatically net profit or cash flow. Comparisons become misleading when one location includes insurance and field labor while another excludes them.

Use the metric to investigate stations or vehicle types that generate substantial revenue but consume disproportionate operating resources. A heavily used vehicle may contribute less than a quieter one if it needs frequent recovery, repairs or expensive repositioning.

4. Asset payback period: when does a vehicle recover its investment?

Asset payback period estimates how long a vehicle takes to generate enough cash contribution to recover its upfront investment. It connects daily operating performance with fleet purchasing decisions.

Simple formula: upfront vehicle investment ÷ expected monthly vehicle cash contribution.

If a vehicle costs €1,500 to acquire and prepare for service, and contributes €125 a month in cash after the defined operating costs, simple payback is 12 months.

Include the upfront costs needed to put the vehicle into operation, such as hardware installation and initial preparation. Use cash contribution before depreciation, rather than revenue or an accounting margin that already deducts depreciation. Assess financing and lease payments separately so costs are not counted twice.

The simple formula assumes stable monthly contribution. For seasonal fleets, use cumulative monthly cash contributions and identify when they recover the initial outlay. Model downtime and expected repairs rather than extrapolating the strongest summer month across the year.

Compare payback with expected service life and the period over which you can operate the asset. Recovering an investment over several years means little if a permit, contract or vehicle condition makes that operating horizon uncertain.

5. Fleet availability: how much capacity is operationally ready?

Fleet availability should help explain why vehicles cannot serve customers. For this framework, measure service-ready capacity: time when a vehicle is either ready for rental or already serving a rental. This separates operational downtime from productive use.

Formula: service-ready vehicle-hours ÷ in-scope fleet vehicle-hours × 100.

A vehicle being rented is working, even though another customer cannot book it. Keep that distinct from immediate bookable availability, which measures the supply a customer can access right now. Label both clearly if your dashboard reports them.

If 10 of 100 vehicles spend an entire day unavailable for operational reasons, while the other 90 remain service-ready, service-ready availability is 90%. Segment the unavailable time into charging, maintenance, cleaning, connectivity issues and other relevant causes.

Wunder Mobility’s operational KPI documentation distinguishes available vehicles, low-battery vehicles, offline vehicles, vehicles in service and active rentals. These states provide useful diagnostic detail; they should not be collapsed into one unexplained unavailable count.

Review duration as well as occurrence. Five vehicles awaiting repair for a week can be more damaging than ten vehicles charging briefly. For station-based fleets, also inspect local bookable supply: a service-ready vehicle elsewhere in the city cannot satisfy a customer waiting at an empty station.

6. Operational cost per trip: what does it cost to deliver a rental?

Operational cost per trip measures the fleet operating expenditure required to support each completed rental. It helps identify whether growth is making operations more efficient or adding work faster than it adds rides.

Formula: defined fleet operating costs ÷ completed trips.

If an operator spends €30,000 on the chosen operating cost pool and completes 15,000 trips, operational cost per trip is €2. Use a consistent cost pool and reconcile it with the costs used in contribution reporting.

Break the figure into activities your team can influence, such as charging, repositioning, cleaning and maintenance. A rising average could come from more repair work, less efficient routes or fewer trips sharing the same fixed workload. Those causes require different responses.

Avoid comparing a short scooter ride directly with a full-day car rental. Segment by service and vehicle type, and review cost per rental-hour where trip duration materially changes the workload or revenue opportunity.

Cutting activity is not the same as improving efficiency. Fewer battery swaps may reduce costs while leaving vehicles unable to serve demand. Evaluate an operational change against availability and contribution margin, not just the number of tasks avoided.

7. Rider retention: do customers return after their first trip?

Rider retention measures whether a defined group of customers continues using the service. Cohort-based tracking shows whether acquisition produces recurring demand or a stream of one-time discounted rentals.

Formula: riders in a defined cohort who complete a rental in a later window ÷ all riders in that cohort × 100.

For example, define a cohort as everyone completing their first rental in January. If 250 of those 1,000 riders complete another rental in February, next-calendar-month retention is 25%.

Keep that definition distinct from the percentage who repeat within 30 days of their first ride. Both can be useful, but they use different observation windows. Only compare cohorts after each has had enough time to complete the window.

Segment retention by city, acquisition channel, first-ride promotion and customer type. Commuters, tourists and occasional carsharing users have different reasons to return. A single fleet-wide percentage can obscure those differences.

Review repeat usage alongside the contribution those rides generate. A voucher campaign may bring customers back while reducing margin. Wunder’s customer lifecycle guidance likewise connects retention with trips per user, revenue per active user and fleet utilization. The operating question is whether returning riders create sustainable demand for your fleet.

8. Station stockout rate: where are customers finding no vehicle?

Station stockout rate measures the share of observed opening time when a station has no bookable vehicles. It shows supply gaps that a fleet-wide utilization figure cannot reveal. For free-floating services, apply the same logic to clearly defined operating zones.

Formula: observed open-station minutes with zero bookable vehicles ÷ total observed open-station minutes × 100.

Suppose a station is monitored from 18:00 to 21:00 on Friday and has no bookable vehicles for 45 of those 180 minutes. Its stockout rate for that window is 25%. Compare the same window across several Fridays to distinguish a recurring gap from a single unusual evening.

Use timestamped availability records, station opening hours and booking eligibility, rather than counting vehicles physically present. A vehicle can be at a station but unavailable to rent. Make this distinction explicit through available-vehicle counts and a separate flag indicating whether a station permits rentals.

An empty station is evidence of missing supply, not proof of lost demand. Where available, compare stockout windows with location-specific searches returning no bookable options, failed reservation attempts attributable to availability, or customer requests. Deduplicate repeated attempts and report them as observed demand signals, not automatically as lost trips. Exclude missing telemetry from measured time and report data coverage separately.

To identify underperforming branches, combine stockout rate with revenue per vehicle per day, utilization and contribution margin. Compare similar service models and operating hours. A high-revenue branch with repeated stockouts may need more capacity; a low-revenue branch with idle, bookable vehicles needs a different investigation. Raw branch revenue alone favors larger fleets and does not explain performance.

9. Invoice collection rate: how much billed revenue has actually been paid?

Invoice collection rate measures the share of a defined set of billed invoices that has been collected by a stated date. It exposes payment gaps that rental revenue and trip counts cannot show.

Formula: payments allocated to an invoice cohort by the reporting cutoff ÷ adjusted billed amount of that cohort × 100.

For example, take €10,000 of invoices issued in October, after applicable credit adjustments. If €9,000 has been collected against those invoices by 30 November, the cohort’s collection rate is 90%, with €1,000 still outstanding. If €300 of that outstanding balance is disputed, show it as part of the €1,000, not an additional amount.

Compare invoice cohorts at the same age, or use cohorts grouped by payment due date, to track the collection trend. An invoice issued yesterday has had less time to be paid than one issued six weeks ago. State the cutoff, use the same currency and VAT basis on both sides, and account consistently for credits, refunds and payment reversals. Keep write-offs visible rather than silently removing unpaid debt to improve the percentage.

For a monthly billing view, show amounts invoiced this month, payments collected this month and the outstanding balance separately. Payments received this month may settle older invoices, so dividing all monthly receipts by all newly billed invoices does not measure collection of the same revenue.

Drill into outstanding balances by due date, branch, customer type and payment status. Separate invoices not yet due from overdue balances, failed payments, payments requiring customer action and disputed amounts. Review disputes by invoice, amount, age and reason to identify what needs resolution. A dispute is not automatically a permanent loss.

Wunder Mobility’s payments and invoices documentation distinguishes invoice states from linked payments and payment attempts, including failed and disputed states. That distinction matters: an invoice status alone does not establish payment success. Use the underlying payment records to understand why billed revenue has not yet been collected.

Use the nine KPIs together

Start with the relationship between the metrics. If revenue per vehicle falls, check whether utilization declined or realized revenue per rental-hour changed. If utilization falls, inspect operational downtime and local supply before assuming demand has weakened. If rides and revenue rise while contribution shrinks, investigate discounts, customer mix and operating costs.

Review station stockouts alongside local demand signals before relocating vehicles. If billed revenue rises while collection weakens, inspect invoice cohorts and payment exceptions before treating the increase as cash available for fleet investment.

Use daily operational views for exceptions, weekly reviews for patterns and monthly financial reviews for margin, collection and payback assumptions. Compare equivalent periods and separate new locations from mature ones. Set targets against your own cost structure, service obligations and seasonal demand rather than importing an industry average with an unknown definition.

The nine metrics are a management core, not a substitute for safety, regulatory or contract reporting. Add the measures your city agreements and service model require.

Get from a KPI to the detail behind it with Wunder Analytics & AI

Shared mobility operations move quickly: a revenue dip or an empty station needs investigation while there is still time to act. Wunder Mobility Analytics & AI brings interactive dashboards and plain-language questions into the back office your team already uses. Review fleet health, station and branch performance, and monetization, then filter the dashboards or ask for a more specific breakdown of your operating data. Wunder AI returns factual answers with charts, using shared KPI definitions across answers and dashboards. Your team can get to the relevant number quickly and drill deeper into the vehicles, locations and periods behind it.