Ask an owner what their deliveries cost and you’ll usually get the fuel number. Sometimes the driver’s hourly rate. Almost never the real figure, because most of what in-house delivery costs never appears on a line item called “delivery.”
This post is the accounting exercise: every cost an in-house delivery operation actually carries, including the five or six that get left out, and how to turn them into a defensible cost per completed drop. It’s not an argument for outsourcing: that’s a separate decision, laid out in outsourcing delivery services. You can’t make that decision honestly without this number first.
Why it matters more than it used to: last-mile delivery has risen from roughly 41% of total shipping cost in 2018 to about 53% today, and the average standard delivery costs around $10.10 to perform while customers are typically charged $8.08 (SmartRoutes last-mile statistics). Delivery is now one of the largest controllable costs in a local business, and plenty of operations are subsidising it without knowing.
The Bottom Line
- Drivers are 50–60% of last-mile expense, and the fully loaded cost of a driver is well above the hourly rate.
- Vehicle maintenance and insurance alone run roughly $5,500–$7,200 a year per food-delivery vehicle, before fuel and before the payment.
- Idle capacity is the largest hidden cost: a van and driver kept for December volume are paid for in February too.
- Failed deliveries average about $17.20 each, and they’re a cost line, not an accident.
Lower your delivery costs by 23%
How we reduce costs:
- No delivery vehicle expenses
- Optimized local routes
- Pay-per-delivery model
- Average 23% delivery cost reduction
Why in-house delivery costs hide so well
Three accounting habits do most of the damage.
The first is that delivery costs are scattered. Fuel sits in vehicle expenses, the driver sits in payroll, the insurance sits in insurance, and nobody adds them up against a delivery count.
The second is that the most expensive inputs aren’t invoiced. Nobody bills you for the two hours a week you spend routing, or for the paid driver hours spent waiting for the last tray.
The third is denominator trouble. Even when the costs get totalled, they’re often divided by deliveries attempted rather than deliveries completed, which understates the number.
Hidden cost 1: the fully loaded cost of a driver
Market rates for delivery drivers sit around $18.45 an hour nationally (ZipRecruiter, September 2026), and local delivery driver roles average closer to $23 an hour (Salary.com).
The rate is the start. Add payroll taxes, workers’ compensation, recruitment and onboarding, training, paid leave, cover when someone’s sick, and the overtime a late route generates. Add turnover: delivery roles churn, and every departure costs a hiring cycle.
Then add the hours that aren’t driving. Loading, waiting for product, returning to base, and the dead time between a route finishing and a shift ending are all paid.
Drivers are 50–60% of total last-mile expense in most operations (SmartRoutes). Getting this line wrong gets the whole exercise wrong.
Hidden cost 2: the vehicle after you’ve bought it
The purchase or lease payment is the visible part. What follows is not.
One 2026 fleet-budgeting guide models a single food-delivery vehicle at roughly $5,500 to $7,200 a year in maintenance and insurance alone, before fuel and before the vehicle payment (Calcix).
Then the rest of it:
- Fuel, which moves with prices you don’t control.
- Depreciation, which is real money even though it never leaves the account.
- Tyres, brakes and servicing on a vehicle doing stop-start urban miles, which wears faster than highway driving.
- Parking, tolls and tickets. In a dense city these are routine, not exceptional.
- Downtime. The week the van is in the shop still has deliveries in it, and covering them costs more than usual.
- Specialist equipment: refrigeration, insulated boxes, racking, a liftgate.
If your orders have outgrown a car or a van, the vehicle question changes shape entirely; box truck delivery service covers what pallet-scale deliveries need.
Hidden cost 3: idle capacity, the big one
This is the cost nobody budgets and the one that usually decides the answer.
Delivery demand for food, floral, catering and wholesale businesses is spiky. Saturday is not Tuesday. December is not February. Valentine’s week is not the rest of the month.
A fleet has to be sized for the peak, or it fails at the peak. Which means for most of the year you’re paying for vehicles and driver hours you don’t need. That’s the inherent cost of owning capacity for a demand curve that moves.
Quantify it: count the driver hours paid in your quietest month and the deliveries completed in it. The cost per drop in a slow month is often two or three times the cost in a busy one. Both are real, and the annual average is the honest figure.
Hidden cost 4: your own management time
Somebody plans the route, re-plans it when an order changes, answers the customer asking where the driver is, handles the breakdown, and covers the no-show.
If that somebody is the owner or a manager, price those hours at what their time is actually worth, not at a driver’s rate. Two hours a day of owner time spent on delivery logistics is the most expensive line in this entire exercise, and it’s invisible in every P&L.
Routing is also a skill with a measurable payoff: properly sequenced routes cut fuel use by something in the 10–25% range depending on how loose the current planning is (NextBillion.ai). If you’re keeping delivery in house, that skill is worth building: efficient delivery planning is where to start.
Hidden cost 5: failed deliveries and redelivery
Every failed attempt is a second trip, a support conversation, and sometimes a refund.
A Loqate study of domestic first-time delivery failures put the average cost at $17.20 per failed order (via PR Newswire). In the US around 97.2% of first attempts succeed, and leading retail operations treat a 90%-plus first-attempt rate as their benchmark (SmartRoutes delivery success rates).
Work out your own first-attempt rate and multiply the failures by a realistic recovery cost. For most in-house operations it’s a bigger number than expected, because an in-house driver with a full afternoon can’t easily go back.
Hidden cost 6: risk, spoilage and liability
Low-probability, high-consequence costs belong in the model as an expected value, not as an afterthought.
Commercial auto liability for an at-fault accident involving your branded vehicle and your employee. Spoiled product on a hot afternoon when a route ran long. Damaged goods with no proof of condition at handover. An injury claim from loading.
None of these happen monthly. All of them are yours.
How to calculate your true cost per delivery
Take one normal month and total these:
| Cost line | Where to find it |
|---|---|
| Vehicle payment or lease | Finance schedule |
| Maintenance, tyres, servicing | Vehicle expenses, annualised ÷ 12 |
| Commercial insurance | Insurance, monthly share |
| Fuel | Fuel receipts for delivery mileage only |
| Driver wages, fully loaded | Rate × hours, plus taxes, comp and benefits |
| Paid non-driving hours | Loading, waiting, returning |
| Idle capacity | Paid hours not spent on a delivery |
| Management time | Hours × what that person’s time is worth |
| Failed deliveries | Failures × recovery cost |
| Equipment and software | Insulation, racking, routing tools |
| Parking, tolls, tickets | Expenses |
Divide the total by completed deliveries that month. Then run it again for your quietest month. The spread between those two numbers is your idle-capacity problem in one figure.
That cost per completed delivery is the only number that makes a courier quote comparable. The decision it feeds (keep driving or hand it over) is in outsourcing delivery services, and the quality bar you should hold any provider to is in what makes a good courier service.
Frequently asked questions
What is a realistic cost per delivery for a small business?
It varies too much with distance, stop density and wages to quote a single figure, which is why the calculation above matters. As a reference point, the average standard delivery costs about $10.10 to perform industry-wide while customers are typically charged $8.08.
Is in-house delivery ever cheaper than a courier?
Yes: at high, steady volume where a vehicle stays busy, and in operations where the delivery itself is part of the product. The problem is that most businesses assume they’re in that group without running the numbers.
What’s the biggest hidden cost of running your own deliveries?
Idle capacity, usually. A fleet must be sized for the peak, so most of the year you pay for vehicles and hours you aren’t using. Management time is a close second because it never appears in the accounts.
How do I compare my in-house cost to a courier quote?
Compare cost per completed delivery over a full month, including surcharges on the courier side and idle capacity, management time and failures on yours. Rate cards alone will mislead you in both directions.
Run the number before the next vehicle
The moment to do this exercise is before you buy the second van, not after. That purchase locks in fixed cost for years against a demand curve you can’t predict.
If the number comes back higher than you expected, that’s useful rather than alarming. It’s the first time delivery has been visible. What you do with it, including the growth case for buying delivery capacity instead of owning it, is covered in the benefits of courier services for small businesses. And if you’re building the shortlist, start with how to choose a courier service for your business.
