How to calculate what a delivery actually costs you

Ask anyone who runs deliveries what a single drop costs and the answer is usually a guess. "Well, fuel, roughly." The problem is not a lack of interest. It is that the cost of a delivery is scattered across places nobody looks at together: this week's fuel, a service three months ago, the hour lost hunting for a badly written address.

When that sum is never done, two things follow. First, the delivery fee gets set by feel, almost always too low. Second, and more quietly, you end up accepting zones or customers that actually lose money, and the business covers the gap with the profitable ones without anyone noticing.

The four costs you need to add up

1. Fuel

The one everybody calculates, and usually calculates wrong. Dividing what you refill each week by the number of deliveries does not work, because that fuel also covers trips that were not deliveries.

What you need is the vehicle's consumption per 100 kilometres and the kilometres you actually drove on the route. If your van uses 12 litres per 100 kilometres and you covered 80 kilometres, you burned 9.6 litres. Multiply by the price per litre and you have the fuel cost for that day.

2. The driver's time

This is where the sum starts getting uncomfortable. The cost is not just the wage: it is the wage plus employment costs, divided by the hours that person genuinely works.

And delivery time is not only driving. It is loading the vehicle, ringing the bell and waiting, taking payment, making change, dealing with the customer who wants to chat. In most urban routes, stationary time is somewhere between 40 and 60 per cent of the day.

If you drive it yourself, still put a number on it. Use what you would pay somebody to run that route.

3. The vehicle, beyond fuel

This is the cost that almost never shows up, because it is not paid daily. A delivery vehicle wears out: tyres, servicing, brakes, insurance, tax, and above all depreciation.

The simple way to turn that into a per-kilometre figure is to add everything you spent on the vehicle over the last year, excluding fuel, and divide it by the kilometres driven in that year. That gives you a cost per kilometre to multiply by the day's route.

4. What gets lost

Failed drops because nobody was home. Returnable containers that never came back. Stock damaged in transit. Extra trips caused by a wrong address.

This block is the hardest to estimate and varies the most. If you have never measured it, start by counting for one month how many deliveries failed out of the total.

The formula

Once you have the four blocks for the day:

Cost per delivery = (fuel + time + vehicle + losses) / completed deliveries

The important detail is in the divisor: completed deliveries, not attempted ones. The failed drops are already counted above as cost.

A worked example

A day with 25 completed deliveries and 3 failed, 80 kilometres driven, one driver on an 8 hour shift:

Item	Calculation	Total
Fuel	9.6 litres	14,400
Driver	8 hours loaded	32,000
Vehicle	80 km at 90 per km	7,200
Losses	2 containers + 1 revisit	4,400
Day total		58,000
Cost per delivery	58,000 / 25	2,320

The number that matters is not the total. It is the 2,320. If you are charging 1,500 for delivery, every drop costs you 820 more than it brings in.

What to do with the number

Once you have it, it drives three concrete decisions.

Setting the delivery fee. With the real cost on the table, the argument stops being whether delivery is expensive and becomes how much margin you want above that floor.

Setting a minimum order. If a delivery costs you 2,320, a 3,000 order makes no sense. That is what minimum order values are for.

Deciding on zones. Cost per delivery is not the same everywhere. A zone with scattered customers can cost twice what a dense one does. With the figure worked out per zone, you can see which to keep, which to price differently and which to drop.

Where the data comes from

All of this depends on having a record of what happened: how many deliveries completed, how many failed, what route was driven, which containers came back. If that lives in the driver's memory or on loose notes, the calculation gets done once, with effort, and never again.

When orders are recorded as they arrive and deliveries are marked as they close, the calculation stops being a project and becomes a query. That is when the number gets genuinely useful: when you can look at it every month without sitting down to reconstruct it.