Rental Car Unit Economics: ROI, Payback and Margin

A rental fleet is not one business — it’s twenty small ones wearing a trench coat. Each car has its own revenue, its own costs, its own payback clock and its own right answer to “should I keep you another year?” Operators who only look at the blended bank balance miss which cars are carrying the fleet and which are quietly losing money.

This guide builds the unit economics of one car from scratch, with Australian numbers, and covers the two places operators most often fool themselves: what counts as revenue, and when to let a car go.

What are the unit economics of one rental car?

One car’s economics reduce to a single line: weekly rate × occupied weeks, minus what the car costs you to own and run.

The revenue side has only two levers. The rate is set by the market and your costs (covered in pricing weekly rentals); occupied weeks are set by your fleet utilization — gaps between renters, weeks lost to the workshop, and weeks occupied but unpaid. A car at $330/week has a theoretical ceiling of $17,160 a year; a realistic well-run figure is 48 occupied weeks, or about $15,840.

The cost side is where optimism goes to die. Using the AAA Transport Affordability Index and Australian market figures: insurance runs about $52 a week for a typical car, servicing and tyres about $35 a week — and rideshare duty is harder than that average, because a full-time driver covers 50,000–60,000 km a year versus roughly 11,000 km for a private car. Rego and CTP add roughly $1,500 a year depending on state. Payment processing takes its cut of every single week — about $6 per $330 card debit, less than half that on PayTo (see cutting payment fees with PayTo). And depreciation — the cost nobody sees until they sell — takes 40–50% of an Australian car’s value by year three (Canstar), faster at rideshare mileage.

What does a $25,000 car actually return?

Here is the full year-one ledger for a $25,000 rideshare-ready car at $330/week and 48 occupied weeks. Figures are illustrative but each line is anchored to the Australian sources above:

Three numbers fall out of that table:

1. Cash payback: ~31 months. $25,000 ÷ $9,550 cash margin per year. That’s how long until the car has returned the money you put in. A widely-used car rental financial model (Financial Models Lab) puts average payback around 41 months — the difference is mostly utilization and the fact that this ledger excludes business overhead.
2. Simple ROI: ~22%. $5,550 economic profit on $25,000 invested. This is the number to compare against other uses of the capital — and the honest one, because it charges the car for its own depreciation.
3. The overhead gap. Industry-wide, overhead consumes 30–45% of gross revenue and net margins land at 5–15%. Per-car economics always look better than business economics; the difference is your time, premises, marketing and admin, spread across the fleet. A 20-car fleet carries overhead better than a 5-car fleet — one reason scale genuinely helps.

Sensitivity is brutal and worth internalising: every occupied week is worth $330, so the gap between 44 and 48 occupied weeks is $1,320 — a quarter of the year’s economic profit. One renter who stops paying for a month costs about the same. On thin per-unit margins, collections and renewals aren’t back-office chores; they’re the profit.

What actually counts as revenue?

Only the rent. Everything else that flows through your account is either someone else’s money or money you’re holding temporarily — and counting it as income is the most common way operators overstate their returns:

• Recurring rent — yes. The weekly payment for the use of the car. This is the only line that belongs in revenue, ROI or payback.
• Deposits — no. A deposit is refundable security. You hold it; you don’t earn it. It goes back at the end of the rental minus documented deductions (how to handle deposits properly). A fleet that books deposits as income shows a great year right up until ten rentals end in the same month.
• Tolls and fines — no. These are pass-throughs: you pay the toll authority, you recover the cost from the renter. Net zero. Only an admin fee you add on top is yours (recovering tolls and fines covers the mechanics).

This stops being pedantry the moment anyone else relies on your numbers. If you fund cars with investor capital and pay a share of “revenue”, a definition that includes deposits and toll recoveries pays investors money that was never earned — and unwinding that conversation is far worse than defining revenue correctly on day one. It’s also why Carz enforces the distinction in the ledger itself: rent, deposits and pass-throughs are separate payment types, and investor reporting only ever sums the rent.

When should a car leave the fleet?

When its forward return drops below what a replacement would earn — and that point arrives earlier than sentiment suggests. Age moves every line of the unit ledger the wrong way at once:

• Maintenance climbs. A rideshare car crosses 150,000 km around year three. Brakes, suspension and transmission bills arrive together, and each workshop visit also costs occupied weeks.
• The rate falls. Renters pay less for a tired car; Australian weekly rideshare rates run from about $165 for older stock to $339 for newer full-size cars (Finder), and your car slides down that curve.
• Depreciation flattens. The one line that improves: by year three the car has already lost 40–50% of its value, so the remaining downside is smaller.

The practical discipline: keep a per-car ledger, review trailing-12-month economic profit per car quarterly, and sell when a car’s trend is heading below your fleet average with a major service on the horizon. Sell before the big bill, not after — the market doesn’t pay you back for a fresh timing chain. Doing this at all requires per-car records rather than one blended bank account; that’s the operational layer covered in the Carz feature set, but a disciplined spreadsheet per car works at small scale too (until it doesn’t — see the real cost of running a fleet on spreadsheets).

The bottom line

One well-run $25,000 car returns roughly $9,500 in cash and $5,500 in true economic profit a year, pays itself back in about 31 months, and earns a ~22% simple ROI. Every one of those numbers is fragile against the same three enemies: idle weeks, unpaid weeks and untracked costs. Run the ledger per car, count only rent as revenue, and let the cars that fall below the line fund the ones that replace them. The fleet is only ever as good as the units in it.