FleetFlow Logo
EcosystemBlogPricing
Get Demo
FleetFlow

Fleet management and dispatch software trusted by 1,000+ operators across 20+ countries. Admin-controlled dispatch, live GPS tracking, and white-label apps, all in one platform.

A product of Hyvikk Solutions

Book Free Demo

Company

  • About Us
  • Pricing
  • Blog
  • Ecosystem
  • Book a Demo
  • Contact Support

Solutions

  • Taxi Fleet Software
  • NEMT Transport Software
  • Corporate Shuttle Management
  • School Transport Software
  • Manual Dispatch System
  • Live GPS Vehicle Tracking

Features

  • All Features
  • Admin Dashboard
  • Driver App
  • Customer App
  • Driver Management
  • Reports and Analytics

Get in Touch

Emailhelp@hyvikk.com

India · Head Office
309, Swara Parklane, Atabhai Chowk, Bhavnagar, Gujarat

Canada · Branch Office
45 Tinton Crescent, Etobicoke, ON M9V 2H9

© 2026 Hyvikk Solutions. All rights reserved. FleetFlow is a product of Hyvikk Solutions.

Privacy PolicyTerms of ServiceRefund PolicyCookie PolicySitemap
Back to Blog
Taxi and private-hire fleet KPIs guide covering revenue per vehicle hour, utilization, deadhead, driver performance, service quality, cost and compliance metrics with formulas and benchmarks for 2026

Taxi and Private-Hire Fleet KPIs: The Metrics That Actually Decide Profit

Harsh Rathi
Jun 26, 2026 • 17 min read

Taxi and Private-Hire Fleet KPIs: The Metrics That Actually Decide Profit

Audio Guide

0:00 / 0:00
0:00

Most fleet management KPI guides are written for trucking.

So they teach you to obsess over the wrong number.

Cost per mile, total cost of ownership, preventive-maintenance compliance. All real metrics. All built for a business that moves freight in heavy trucks across long distances. If you run a taxi or private-hire fleet, not one of them is the number that decides whether you make money.

I work on FleetFlow, and we have been building software for taxi and minicab operators since 2015, back when the product was still called Fleet Manager and ran on CodeCanyon. Across more than 700 operators in over 25 countries, the same mistake keeps surfacing in the data. The owner watches cost per mile like a hawk, while the thing actually draining the business, empty vehicle hours, never makes it onto the dashboard at all.

Here is the number that should be there instead: revenue per vehicle hour. How much each car earns for every hour it sits on shift. It is a function of utilization, deadhead, driver behaviour, and service quality, and once you get it right, most of the cost metrics quietly take care of themselves.

Quick definition first, because the listicles skip it. A metric is any number you can measure. A KPI is a metric tied to a goal you actually care about. Every KPI is a metric. The reverse is where most dashboards go to die.

One more thing before the list, because it explains why this guide had to exist at all. The fleet-management software market is worth somewhere between $30 and $38 billion in 2026, depending on which analyst you believe, growing in the mid-to-high teens. Read any of those reports, though, and you find them sliced by logistics, construction, and e-commerce. Taxi and private hire never get their own row. The market data is trucking-shaped, the KPI guides are trucking-shaped, and that is precisely the gap this post fills.

So this is not a list of 25 metrics. It is the shortlist, grouped by what each number controls, with a plain formula, an honest benchmark, and the four or five worth watching every single day.


Last reviewed: June 2026. The utilization, deadhead, earnings, and market figures here are sourced and linked inline. Where a study is more than a couple of years old, I have flagged it in the text, because a stale benchmark quoted as current is worse than no benchmark. Treat every range as a reference point, not a target, and baseline your own fleet first.

How many KPIs should you actually track?

Fewer than you would think, and far fewer than the listicles sell you.

Run three to five north-star KPIs daily, and keep everything else as a drill-down you open only when a north-star number twitches. Twenty-five metrics on a wall is not visibility. It is wallpaper, and by Wednesday nobody is reading it.

A KPI has to earn its spot. The test is brutal and simple: if the number moving would not change a single thing you do, it is not a KPI. It is decoration.

Now the part that matters more than the shortlist itself. Before you chase anyone else's benchmark, baseline your own fleet. Three to six months of your own numbers beats every industry average going, because your city, your particular mix of airport runs and school contracts and corporate accounts, and the way you engage drivers are not the average. They are yours.

And be hard on the benchmarks you do borrow. A whole tier of taxi-KPI content online is template spam built around fictional companies, and the numbers in it are not just optimistic, they are impossible. I have sat and read guides quoting 600 percent fleet utilization and 820 percent gross margins without a hint of embarrassment. You cannot use more than 100 percent of a vehicle's time. When a benchmark breaks basic arithmetic, it is filler dressed up as data. Trust the numbers that carry a named, real source, and treat the rest as noise.


Revenue and asset-productivity KPIs (start here)

If you only had a weekend to build a dashboard, you would build this section and ignore the rest until Monday. This is where private-hire profit is actually won or lost.

Revenue per vehicle hour

The most important number you are probably not tracking.

Revenue per vehicle hour = total fleet revenue ÷ total hours vehicles were on shift.

You can slice the same idea per vehicle per day or per week, but per active hour is the cleanest read of the lot. And here is the contrast that decides everything. A freight truck earns by covering distance, so cost per mile is genuinely its profit lever. A taxi earns by being booked while it is out. A car can rack up 200 miles in a shift and bring home almost nothing, because most of those miles were empty. Revenue per vehicle hour catches that in one figure. Cost per mile hides it completely.

Vehicle utilization rate

Utilization is the share of available shift time a vehicle actually spends earning. Easy to say, slippery to measure, because two other industries borrowed the word and meant different things by it. Trucking calls a vehicle utilized when it is simply not broken down. Car rental measures days rented over days available. Neither maps to your business. For private hire, utilization has to mean shift time spent earning, because that is the only version tied to the money.

Utilization rate = earning time ÷ total available shift time × 100.

Now the honest benchmark, and brace yourself. Spot-market ride utilization usually runs 30 to 60 percent. Independent academic work, counting both time carrying a passenger and time driving to one, found it peaks near 50 percent even at peak demand. Sit with that for a second. At the busiest hour of the week, with demand surging and surge pricing live, roughly half your fleet's time is still unpaid. That is not slack you coach away with a team pep talk. It is structural to spot-market dispatch, and it is the single biggest lever in this entire post. A managed fleet leaning on pre-bookings and corporate contracts should comfortably beat raw spot-market numbers, but do not treat 50 percent as either a target or a ceiling. Baseline yours, then climb from there.

Deadhead ratio (empty time)

Deadhead is utilization's shadow: the share of your miles or minutes spent carrying nobody. Pure cost, zero revenue, and the maintenance-first guides never even name it.

Deadhead ratio = empty miles (or minutes) ÷ total miles (or minutes) × 100.

The scale is bigger than most owners guess. A conservative estimate puts around 40 percent of ride-hailing miles as travelled empty. Average occupancy sits near 1.4 passengers a ride, and drops to about 0.8 once you fold in the repositioning miles. Idle time clumps in predictable spots too: one city study clocked 22 to 27 minutes of driver idle time at the airport, and up to half an hour mid-afternoon in the quiet zones.

Two honest caveats before you panic. This is ride-hailing data, and a managed fleet that pairs return trips and runs pre-bookings should do better than a lone app driver chasing pings across town. But the lesson does not move: every empty mile is fuel burned, brakes worn, and a driver paid, all for nothing. That airport figure is exactly why dispatch logic earns its keep, and it is the metric the trucking guides leave out entirely.

Average fare and trips per shift

Two quick reads sit underneath that headline revenue number. Average fare, which is total fare revenue divided by completed trips, tells you about pricing power and trip mix. Trips per vehicle per shift, completed trips divided by vehicle shifts, tells you about demand and how well dispatch is filling the day. When revenue per vehicle hour sags, these two between them tell you whether the problem is price or volume. That is the entire reason a drill-down exists.

Revenue KPIFormulaHonest benchmark
Revenue per vehicle hourTotal revenue ÷ vehicle shift hoursBaseline your own; this is the north star
Utilization rateEarning time ÷ available shift time × 100Spot-market 30-60%, peaks ~50%
Deadhead ratioEmpty miles ÷ total miles × 100~40% of ride-hail miles can be empty
Average fareFare revenue ÷ completed tripsSet from your own market
Trips per vehicle per shiftCompleted trips ÷ vehicle shiftsTrack the trend, not a fixed target

Driver performance and retention KPIs

Every revenue number above runs on drivers. Lose them and the dashboard goes dark, because a parked car with nobody to drive it earns exactly zero. What follows is the summary view. The full playbooks live in driver behavior analytics and reducing driver turnover, so I will keep each one tight here.

Acceptance rate, jobs accepted over jobs offered, is your earliest warning light. When it starts to slide, drivers are either disengaging or the job mix has turned unattractive, and that slide usually begins weeks before anyone hands in notice. Cancellation rate is only useful once you split it. Driver-initiated cancellations point back at allocation and earnings friction. Passenger-initiated ones point at wait time and ETA accuracy. Lumped into one figure they tell you nothing at all.

Completed trips per active driver

This is a number that quietly lies to you if you measure it the obvious way. It is completed trips divided by active drivers, and the word doing all the work is active. A big share of drivers are part-time. One Beijing dataset found about half of them drove under 10 hours a week, and while I would treat that exact share as local colour rather than gospel, the structural point travels everywhere: a roster is not a workforce. Measure per active driver, not per name on the list, or a pile of weekend-only drivers will make a perfectly healthy fleet look broken.

Driver earnings per hour and churn

This is where the driver metrics loop straight back into revenue. Driver earnings are thin. One Denver study put net wages at roughly $5.72 to $10.46 an hour before tax. Now read the date on that figure: 2019. That is seven years old, which in this trade is ancient history. A pandemic, a fare-inflation cycle, the post-2021 surge in driver supply, and round after round of platform commission changes have all landed since. The exact dollar figure is almost certainly stale, and anyone quoting it in 2026 as a current wage is not paying attention.

The pattern underneath it, though, has not shifted an inch. Spot-market driving pays thin and pays unevenly, and thin uneven pay is exactly what makes a good driver pick up the phone when a rival operator calls. That is why earnings per hour and turnover belong on the same screen. Track turnover as drivers who left divided by average active drivers, and watch your first-90-day drop-off most closely, because early churn is both the cheapest to fix and the loudest signal about your onboarding.

Driver safety and behaviour score

Harsh braking, speeding, and cornering rolled into a single score per driver. It feeds insurance cost, vehicle wear, and complaint rate, so it earns its place on the list, but it deserves a full post rather than a paragraph. Go deep in the driver behavior analytics breakdown.

Driver KPIFormulaWhat it tells you
Acceptance rateJobs accepted ÷ jobs offered × 100Driver engagement and job appeal
Cancellation rateCancelled ÷ booked × 100Allocation vs wait-time problems
Trips per active driverCompleted trips ÷ active driversReal productivity, part-time adjusted
Driver turnover rateDrivers left ÷ avg active drivers × 100Retention health and hidden cost

Service quality and demand KPIs

Revenue tells you what already happened. Service quality tells you whether it happens again next month, which for a fleet living on repeat bookings and corporate accounts is the more important question by a distance.

Average wait time, the total time from booking to pickup divided by completed trips, is the number that decides whether a passenger opens your app next time or a competitor's. A rider who waits too long once rarely waits twice. ETA accuracy matters just as much, and sometimes more, because a broken promise stings worse than an honest long wait. Both of them lean entirely on real-time vehicle tracking.

Then there is job fulfillment rate, trips completed over trips requested. It has an evil twin, the unfulfilled-booking rate, which is simply the demand you handed a competitor for free. If fulfillment is low while utilization is also low, stop blaming the market. Your problem is dispatch, and dispatch is fixable. Airport contracts make this worse, because a missed pickup there is not just a dropped fare, it is a compliance and SLA problem at the same time. The breakdown is in airport taxi dispatch software: what operators actually need. And if fulfillment is low while utilization is high, the constraint is demand outrunning supply, which is a pricing question before it is a dispatch one. That is dynamic and surge pricing for taxis.

Finally, two retention signals. Average customer rating is satisfaction. Repeat-booking rate, returning customers over total customers, is loyalty, and it is the cheaper of the two to grow, because winning back a rider you already have costs a fraction of finding a new one. For corporate accounts it is basically the whole game: repeat rate is just the contract, renewed.


Cost and margin KPIs

Cost matters. It just is not the north star here, and the trucking guides get that backwards. Watch this category monthly, not daily, and for the full cost playbook read how taxi fleet tracking reduces costs.

The private-hire reframe of cost per mile is cost per trip, total operating cost divided by completed trips. You will see the trucking world quote cost-per-mile ranges everywhere, somewhere around $0.23 to $0.61 a mile. I am only citing those so you can recognise the number and walk straight past it. They are heavy-asset figures for vehicles that earn by the mile. Benchmark a taxi fleet against them and you will spend your energy optimising the wrong thing entirely.

Maintenance cost per vehicle and fleet availability, the share of cars actually ready to work, both feed back into utilization, because a vehicle in the workshop is a vehicle earning nothing. The trucking guides chase 95-percent-plus preventive-maintenance compliance. Useful discipline, wrong obsession for a 30-car minicab fleet. Get the routine right with a preventive fleet maintenance checklist and move on.

Energy or fuel cost per trip is where the electric-versus-diesel argument finally shows up in real money, and the savings live in your depot, not the showroom. The full math is in the electric taxi fleet cost and ROI breakdown.

And then the one number that actually tells you whether you are making money: contribution margin per trip, which is average fare minus the variable cost of running that trip. A fleet can be flat-out busy and still bleeding cash if this figure is thin. Busy is not the same as profitable. This is the line that knows the difference.


Compliance and risk KPIs

These do not grow revenue. They stop it from disappearing overnight. One lapsed insurance certificate or one serious incident can cost more than a slow month, and unlike a slow month, it can end the business outright. Three are worth keeping an eye on.

  • Credential currency: the share of drivers and vehicles with valid, in-date documents. Aim for 100 percent and treat anything less as a live risk, because an expired badge silently pulls a driver off the road through no fault of theirs. The country-by-country detail sits in our taxi licence and permit guide.
  • Inspection or MOT compliance: vehicles with a current inspection over total vehicles. Same logic, same 100 percent target. Track expiries weeks out, not the morning they lapse.
  • Accident and complaint rate: incidents over total trips or miles, paired with complaints per thousand trips. Both feed insurance premiums and licence risk, and a rising complaint rate is usually the first visible crack before something bigger gives way.

The thread running through all six categories is the same. The numbers that decide private-hire profit are revenue per vehicle hour, utilization, and deadhead, not cost per mile and PM compliance. Get each vehicle earning more per hour on shift, and the cost figures mostly fall into line behind it. That is the whole argument.


The 5 KPIs to put on your daily dashboard

You do not check 25 numbers with your morning coffee. You check five. Everything else is a drill-down for the morning one of these looks wrong.

Daily KPIWhat it tells youCadence
Revenue per vehicle hourAre your assets actually earningDaily
Utilization rateHow much shift time is wastedDaily
Driver acceptance and retentionAre drivers engaged and stayingDaily / weekly
Average wait timeIs service good enough to keep ridersDaily
Job fulfillment rateHow much demand you are droppingDaily

Cost, margin, and compliance do not need daily eyes on them. They move slowly. Glance at them monthly and you will catch every trend that matters without burying the dashboard in noise.

Here is the honest snag. Tracking all of this by hand works fine at ten cars. Push past fifty and the spreadsheet falls behind, the numbers go stale, and the dashboard you built to make decisions becomes a file nobody trusts. One of our operators in Ghana put it plainly after switching off spreadsheets: it cut their admin time in half. A platform calculates every KPI in this post automatically, from live trip, driver, and vehicle data, on the FleetFlow admin dashboard, so the numbers update themselves while you run the fleet.

See Your Fleet's KPIs Live, Not in a Spreadsheet

Revenue per vehicle hour, utilization, deadhead, acceptance rate, wait time, and fulfillment, all calculated automatically from live data. See your real numbers in one 15-minute walkthrough.

Book a Demo

Frequently Asked Questions

What is the difference between a KPI and a metric?

A metric measures any quantifiable activity, like trips completed or miles driven. A KPI measures progress toward a specific operational goal, like revenue per vehicle hour against a target you set. Every KPI is a metric, but not every metric is a KPI. The skill is picking the few metrics that actually move your business and treating only those as KPIs, so your dashboard stays decision-useful instead of cluttered.

How many fleet KPIs should I track?

Three to five for daily visibility, and the rest as diagnostic drill-downs. Tracking everything equally causes dashboard fatigue and decision paralysis. Pick a north-star set that covers revenue, drivers, and service, review those daily, and pull the deeper cost and compliance metrics monthly. The point of a KPI is to trigger a decision, not to fill a screen.

What is a good vehicle utilization rate for a taxi fleet?

Be careful with this number. Spot-market ride-hailing utilization typically runs 30 to 60 percent, and independent studies show it peaks near 50 percent even at peak demand. A managed fleet with pre-bookings and corporate contracts should aim higher than spot-market work, but ignore the fabricated figures floating around online. Baseline your own utilization over three to six months first, then set a target above your real starting point.

What is the single most important KPI for a private-hire fleet?

Revenue per vehicle hour. It is total fleet revenue divided by the hours your vehicles were on shift. It folds utilization, pricing, and demand into one number, and it answers the only question that really matters: is each vehicle earning while it is out there. Cost per mile, the metric the trucking guides lead with, is secondary for private hire.

How do I track fleet KPIs without a spreadsheet?

A spreadsheet works until your fleet grows past a handful of vehicles, then manual entry breaks and the numbers go stale. A fleet platform like FleetFlow calculates these KPIs automatically from live trip, driver, and vehicle data, so revenue per vehicle hour, utilization, acceptance rate, and wait time update themselves on an admin dashboard instead of in a file you keep forgetting to refresh.


Sources

  • BusinessDojo, Profitability of ride-hailing. Driver utilization typically runs 30 to 60 percent.
  • arXiv, Spatiotemporal characteristics of ride-sourcing operation. Utilization peaks near 50 percent even at peak demand.
  • U.S. DOT, Bureau of Transportation Statistics. Roughly 40 percent of ride-hailing miles travelled without a passenger, occupancy ~1.4 falling to ~0.8 with deadheading.
  • arXiv, Ride-sourcing search frictions and driver productivity. Airport idle time of 22 to 27 minutes.
  • arXiv, Ride-hailing mobility study, Beijing. About half of drivers worked under 10 hours per week. Treat the exact share as one city's data, not a global constant.
  • ScienceDirect, Individual economics of ride-hailing drivers. Net hourly wages of ~$5.72 to $10.46 in a Denver study, 2019. Cited for the pattern, not the dollar amount, which is now seven years old and almost certainly out of date.
  • Global Market Insights, Fleet Management Market. 2026 market-size and CAGR range, cited alongside Fortune Business Insights and Future Market Insights.
  • Oxmaint and Maintainly, fleet KPI guides. The 3 to 5 north-star KPI principle, dashboard-fatigue warning, and baseline-first guidance. Cost-per-mile and PM-compliance benchmarks cited as trucking figures, not taxi targets.

Share this article

On this page

  • How many KPIs should you actually track?
  • Revenue and asset-productivity KPIs (start here)
  • Driver performance and retention KPIs
  • Service quality and demand KPIs
  • Cost and margin KPIs
  • Compliance and risk KPIs
  • The 5 KPIs to put on your daily dashboard
  • Sources

Read Related

Taxi Dispatch Software & FHV Fleet Industry Report 2026: Data, Trends, and Operator Playbook
Industry Reports•May 5, 2026

Taxi Dispatch Software & FHV Fleet Industry Report 2026: Data, Trends, and Operator Playbook

A source-backed 2026 taxi dispatch software and FHV fleet industry report for operators: market benchmarks, taxi fleet management software trends, EV cost math, driver retention, corporate booking, and dispatch system strategy.

Read Article
Electric Taxi Fleet: The Real Cost, Charging, and ROI for Operators
Fleet Management•Jun 15, 2026

Electric Taxi Fleet: The Real Cost, Charging, and ROI for Operators

An electric taxi can cost more than petrol. I have watched it happen. The energy savings are real, but they live in your depot, not the showroom. Here is the operator-grade math: real cost per mile, when EVs do not pay, the charging decision that makes or breaks ROI, payback at taxi mileage, and the 2026 rule changes that rewrote the spreadsheet.

Read Article
How to Manage Multi-Branch Fleets Without Losing Control
Fleet Management•May 1, 2026

How to Manage Multi-Branch Fleets Without Losing Control

Multi-branch fleet operations fall apart without a central system. Here's how to manage multiple locations, drivers, vehicles, and revenue from one dashboard without losing control.

Read Article