Ask what a delivery costs and the answer that comes back is usually a cost per drop built from two lines: the driver’s wages and the diesel, divided by the day’s stops. It feels rigorous because it uses real numbers. The problem is what it leaves out — the vehicle’s standing costs, the warehouse time that put the load together, the office time that processed the paperwork, and the drops that had to be done twice. Price a marginal account, a delivery charge or a minimum order off the two-line number and the decision can easily be wrong by a wide margin. Here is how to build the true figure, with a worked example and a calculator to run your own.
Why the back-of-envelope number misleads
Cost per drop is a full-allocation question wearing a simple disguise. A route consumes five kinds of cost, and the two-line version counts only the first:
- Variable running costs — the costs that scale with mileage: fuel, plus maintenance and tyres where you allocate them per kilometre. (The fixed-versus-variable distinction matters here: pick one home for each cost line and keep it there, or the same tyre bill ends up counted twice.)
- Driver cost — the full employment cost of the hours, not just the headline hourly rate: employer’s NI, pension, holiday cover.
- Vehicle standing (fixed) costs — finance or depreciation, insurance, tax, MOT and compliance. These accrue whether the van moves or not, and each route-day must carry its share.
- Upstream labour — the picking, packing and loading that built the load, and the routing and admin that planned it.
- Failure and rework — failed deliveries redelivered, disputed drops credited, returns hauled back and processed. These don’t appear on any line called “cost”; they appear as extra stops and office hours that dilute every good drop.
The cost per drop method: cost one route-day fully, then divide
Work with one representative route over one day. Annual and monthly costs get divided down to a day; day costs stand as they are; then the total is divided by successful drops — a failed drop is a cost, not an output.
| Cost line | Assumption | Per route-day |
|---|---|---|
| Driver (full employment cost) | 9 hrs at £15.00 fully loaded | £135.00 |
| Fuel | 120 km at £0.16 / km | £19.20 |
| Two-line subtotal | £154.20 | |
| Vehicle standing (fixed) costs | £7,200 / yr (finance, insurance, tax, compliance) over 240 route-days | £30.00 |
| Maintenance & tyres (mileage-allocated) | 120 km at roughly £0.083 / km | £10.00 |
| Pick, pack and load labour | 2.5 warehouse hrs at £13.20 | £33.00 |
| Routing and order admin | 1.5 office hrs at £14.00 | £21.00 |
| Failure and rework | 1 failed drop redelivered + credits processed | £28.90 |
| Full route-day cost | £277.10 |
Now the division. The route was planned for 23 drops; one failed, so 22 succeeded.
The gap is not an accounting curiosity. If a marginal customer’s drop generates £10 of gross margin, the two-line number says the visit is comfortably profitable and the true number says it loses money before anyone has picked up a phone about it.
Three versions of the number — and which decision each one serves
“Cost per drop” is really three metrics wearing one name, and using the wrong one for a decision is how good analysis produces bad calls:
- Wheels-only route cost (driver + variable running + vehicle standing). The fleet manager’s number: it compares routes, vehicles and days on a like-for-like basis, and it is the right lens for routing and fleet decisions.
- Controllable operating cost (wheels-only + pick, load and admin). The operations number: it shows what the whole fulfilment machine spends to put one delivery on a doorstep, and responds to the levers operations actually holds.
- Full operating cost (everything above, including failure and rework). The commercial number — and an essential input to pricing an account, setting a delivery charge or a minimum order. One scope note keeps it honest: this allocates the route and fulfilment operating costs, not the whole business — management, premises, IT, sales and the profit target still have to be recovered on top, so use it alongside margin and overhead recovery, not instead of them.
One more distinction for completeness: the marginal cost of adding one stop to an existing route is lower than any of the three, because the van is already out. That is the right number for “should we squeeze this extra drop in today?” — and the wrong one for deciding whether an account is worth serving at all.
How sensitive is it to drop count?
Very — and this is the cleanest way to see what route density is worth. Holding the illustrative route-day cost at £277.10:
| Successful drops | Full operating cost per drop |
|---|---|
| 18 | £15.39 |
| 22 | £12.60 |
| 26 | £10.66 |
Four extra successful drops on the same route-day moves the number by roughly £2 in either direction — which is why drop density is often one of the strongest levers available. One caveat keeps the table honest: it isolates the density effect by holding the route-day cost constant, and in reality extra drops add some time and mileage of their own. Read it as a sensitivity, not a forecast.
Counting failed deliveries without double counting them
A failed drop creates cost twice — the wasted attempt today and the redelivery later — and the bookkeeping rule is: each day carries only its own events. Today’s failed attempt reduces today’s successful-drop denominator (the attempt consumed route time for nothing); the redelivery is simply one of the stops on whichever day it runs. The “failure and rework” line then holds only the incremental costs failures trigger — the extra office handling, credits processing, returns hauled back — not a second copy of the redelivery’s route time, which its own day already carries. Follow that rule and the numbers add up across the week; break it and failures either vanish or get counted twice.
Run your own numbers
To run your own numbers, use the worked build-up above as the template: swap each cost line for your own figure, add them into a route-day total, and divide by the successful drops that route-day actually achieved.
What actually moves cost per drop
Once the full build-up is visible, the levers become discussable with numbers attached. Failure and rework is often worth examining early, because it creates cost without producing a successful delivery — and its two halves respond to different fixes: planning around delivery windows and keeping the customer informed of the ETA can reduce failed attempts in the first place, while clean electronic proof of delivery strengthens the evidence and exception handling so fewer completed drops turn into disputes and credits. Mileage and hours respond to sequencing — route planning that auto-optimises stops and respects vehicle constraints takes cost out of every drop on the run rather than one drop at a time. Drop density is the third lever, and it needs stating carefully: density improves when the same route time serves more successful stops — tighter geography, fewer failures, better sequencing. Serving a thin account less often does not by itself raise the day’s successful drops; it lowers cost by freeing capacity, and the gain is only banked if that capacity is refilled with denser stops.
Cost per drop is not profit per drop
Which is why minimum order values belong in this conversation but not in this formula. A minimum order raises the revenue and margin each drop carries; it does not change the cost of making the stop. Account decisions need both numbers side by side — the full operating cost of serving the customer at their frequency, and the margin their orders contribute per drop. A customer can clear a healthy margin percentage and still fail the comparison because their drops are small and far apart; another can look thin on margin and pass easily on density. Fixing the frequency, the minimum, or the model (a question that loops back to the van-sales-versus-pre-sales segmentation) is then a commercial choice made with real numbers instead of an average.
The measurement itself is the part software changes most. Because RouteMagic runs order capture, warehouse, routes, deliveries and invoicing on one platform, the ingredients of the calculation — hours, mileage, stops completed versus failed, credits raised — already exist against each route rather than in three systems and a diary, and reporting and analytics can put daily activity, profitability and driver performance in front of the person pricing the next account. Plans are priced per user, not per vehicle — the published pricing makes the platform side of the equation a known quantity.
Conclusion
There is no single cost per drop that serves every decision — and pretending there is causes most of the trouble. Fleet decisions want the wheels-only view, operations wants the controllable view, and commercial decisions need the full operating cost read alongside margin, overhead recovery and the profit target. The method underneath all three is the same and it is deliberately unglamorous: cost one representative route-day completely, keep fixed and variable costs in separate homes so nothing is counted twice, divide by successful drops, and let each day carry only its own events. The two management views worth keeping permanently are your own number tracked over time and the spread between your densest and thinnest routes — both say more than any external benchmark could. A practical next step: run the same method on three representative routes — a dense one, a thin one, a typical one — before changing minimum orders, visit frequency or fleet size. The spread between those three numbers is usually the strategy conversation.
