Ask whether a taxi business is profitable and you will get answers ranging from “I cleared six figures” to “I sold the car and went back to employment” — both from people telling the truth. That spread is the honest answer: this trade has no typical profit, because profit here is decided by a handful of ratios that vary enormously between two operators in the same town.
So instead of quoting an industry average I would not be able to verify, here are the four numbers that actually decide the question, and the arithmetic to run on your own situation before — or after — you commit. Where I use figures below, they are worked examples with stated assumptions, not market statistics; substitute your own numbers, because yours are the only ones that matter.
The number that decides everything: paid miles
A taxi earns nothing while moving without a passenger. The single strongest determinant of profitability is the share of your working hours and miles that are paid — utilisation, in fleet language; “dead miles” is the cost side of the same coin.
Work the example. A driver doing a 10-hour day with the meter or the booking sheet earning £25 an hour while occupied takes £250 on a fully-utilised day — which never happens. At 60% utilisation the same day earns £150. At 40%, £100. Same car, same hours, same rates: the difference between a living and a loss is entirely in the empty gaps.
This is why business model matters more than effort. Street and rank work buys utilisation with waiting; app work buys it with commission; pre-booked work buys it with scheduling — an airport-transfer operator who books tomorrow’s diary tonight knows their utilisation in advance. It is also why the German Mietwagen rule about returning to base empty (the Rückkehrpflicht) is such an economic weight there — regulation can set your dead miles for you.
The overhead that exists before the first fare
A private hire business carries a layer of cost that exists whether or not the phone rings: the operator, driver and vehicle licences, insurance for hire and reward, the vehicle itself, and compliance checks. None of it is optional and most of it renews on a clock.
I am not going to restate licence fees here, because they vary by a factor of several times between neighbouring councils and change most Aprils — the worked figures from two councils’ published schedules are in the UK start-up guide. The point for profitability is different: divide your total annual fixed overhead by your realistic annual jobs before you start. An overhead of £6,000 a year spread over 6,000 jobs is £1 a job; spread over 1,500 jobs it is £4 a job — often the difference between your margin and nothing, and it is set by volume, not by effort.
This is also the honest answer to “how much do taxi business owners make”: the owner’s income is whatever is left after that overhead and after the driver is paid — and if the owner is the driver, it is easy to spend a year confusing wages with profit. Pay yourself a notional driver’s wage on paper first. What remains is the business’s profit. If that number is near zero, you own a job, not a business — which is fine, as long as you priced it knowingly.
Where the margin actually lives
Not all fares are equal, and the profitable operators are the ones who know their mix.
Cash street work is high-volume, low-margin, and utilisation-hungry. App platform work trades commission for demand — you buy utilisation and sell margin. Pre-booked private work — airport runs, chauffeur jobs, school contracts — carries the best margin per mile because it is scheduled, priced in advance and often repeat. Account and corporate work adds invoicing lag but compounds: one won account is hundreds of bookings.
Two margin levers are chronically under-used. The first is repeat share: winning a new customer costs marketing money; a returning one costs nothing, so the share of bookings from existing customers is a direct profit ratio. The second is the overflow you decline — every job you turn away because the diary is full was margin you had already paid to win; subcontracting it to a partner at an agreed cost keeps part of that margin instead of donating the whole job to a competitor.
And one lever people rank too high: fleet growth. Adding a car adds overhead with certainty and revenue only with sufficient demand. A second vehicle at 40% utilisation makes the business less profitable, not more. Grow when the first car is turning work away weekly, not when it merely feels busy.
The costs that decide it quietly
Fuel and the driver are visible. The quiet profit-killers are duller: the unrecorded job that was driven but never invoiced; the account invoice nobody chased for ninety days; the software and phone stack paid for twice; the insurance renewal accepted without requoting; the vehicle downtime week where overhead ran and revenue did not. None of these appears in the “is it profitable” daydream, and together they routinely outweigh the fuel bill.
This is where record-keeping stops being compliance and becomes economics. The same discipline your licence already requires — every booking recorded with its details — is also the only way to know your utilisation, your repeat share and your revenue per route. We build RideDesk, so read this as an interested party’s note: the analytics side of it exists precisely because of this overlap — revenue, top routes, repeat-customer share and subcontracted margin are read straight off the bookings you were obliged to record anyway, and every booking logs which marketing source produced it, so “which work actually pays” is a report rather than a feeling. The free tier’s ten transfers a month covers an owner-driver testing exactly these ratios; an operator running on paper can compute all of them with a spreadsheet and patience — the maths does not care which tool you use, only that the records exist.
The honest verdict
Is a taxi business profitable? Run well and pre-booked-heavy, at genuine volume: yes, sustainably. Run at low utilisation with unexamined overheads: no, and no amount of extra hours fixes arithmetic.
The uncomfortable part is that most founders can compute their answer before buying anything: realistic jobs a week × realistic average fare, minus a driver’s wage for the hours, minus fuel at your dead-mile ratio, minus the annual fixed overhead divided over those jobs. If the result only turns positive at a utilisation you have no plan to achieve, the plan is the problem — and if it turns positive comfortably, the next constraint is demand, which is a visibility problem, not an operations one.
Do that sum on one page, with your own numbers, before anyone sells you anything — including us.
