You know that feeling when you finish a Friday-evening unload three hours from base with no return load, and you do not really care. The truck rolls home empty and somehow it does not sting. Run the same 300 km empty leg at 09:00 on Tuesday in a peak week, and the gut reaction is different, even though the fuel bill, the driver wage, and the toll are nearly identical.
That instinct is correct. Empty kilometers do not have a single price. They have a price that swings between "almost free" and "the most expensive kilometer you will drive this month," and most CPKM models collapse all of it into one flat number that is too low for half your decisions and too high for the other half.
This is the gap that experienced dispatchers patch with gut feel and that newer ones get burned by. It is also where the difference between a fleet running at 14 % margin and one running at 6 % margin tends to live, even with similar costs and similar rates.
Direct cost is the easy half
Every operator with three years and a spreadsheet has a CPKM, and the empty-km variant of it is straightforward:
- Fuel: empty truck uses 8–12 % less than a fully loaded one (so ~25 L/100 km empty vs ~28 L/100 km loaded for a Euro 6 tractor at cruising speed)
- Driver wage: same per hour
- Tolls: same per kilometer for the same axle/emission class
- Depreciation, service, insurance: same per km
Round it to "an empty km costs about 90 % of a loaded km in direct cost." That number goes into the spreadsheet. It is also where most CPKM thinking stops.
The problem is direct cost is roughly half of what an empty kilometer actually takes from you.
The opportunity cost flips the math
The other half is opportunity cost: what those driver-hours and that truck-position could have been doing instead. The same 100 km empty leg carries very different opportunity cost depending on when and where it happens.
Three classes worth naming:
Class 1. Sunk-time empty
The driver is heading toward base, close to the hours-of-service limit, with no realistic chance of picking up another load before mandatory rest. The 200 km back home is empty because nothing else can be done with that vehicle and that driver in that window. Opportunity cost: close to zero. The hours would have been "lost" to rest at the destination instead of rest at home.
This is the cheap empty leg. The real cost is just the direct cost, and arguably less, because returning the truck and driver to base creates scheduling flexibility for the following Monday.
Class 2. Peak-time empty
Tuesday morning, peak week, spot rates 15–25 % above corridor average, multiple alternative loads available within 50 km that you turned down to run this positioning. Direct cost is the same as Class 1. Opportunity cost is everything you did not earn by sending the same driver-hours and truck-km to the better option you ignored.
This is the empty leg that quietly destroys margin. Not because it costs more to drive (it does not) but because of what was sitting on the other end of the choice tree.
Class 3. Positioning as investment
You drive empty toward a region with high backload density and a known rate structure, on the bet that the next loaded leg out of there is worth more than the cost of getting there. Direct cost is paid; the offset is the expected value of the next job.
If 70 % of the time you find a €1.25/km × 350 km load out of the Ruhr within 18 hours of arrival, the expected positioning benefit on a 100 km empty leg is something like:
0.7 × (350 × €0.25 margin) = €61 expected upside
That is a real number to compare against the €90 direct cost of the positioning. The leg is profit-positive in expectation even though it is empty. Most small carriers do not run this math because tracking regional backload probability is a separate job that nothing in their stack does for them.
Same leg, three real prices
Run a 100 km empty positioning under each class. Direct cost is roughly the same in all three:
- Fuel: ~22 L × €1.50 = €33
- Driver: ~1.5 h × €15 = €22.50
- Tolls: ~100 km × €0.18 = €18
- Wear (depreciation + service + insurance): ~€8
- Direct total: ~€81
Now apply the opportunity layer:
| Scenario | Class | Direct | Opportunity layer | Real cost |
|---|---|---|---|---|
| Friday 18:00, heading home, driver near HOS limit | 1 | €81 | €0 (no alternatives) | €81 |
| Tuesday 09:00, peak week, rejected nearby 200 km load @ €1.30 (€0.30 margin × 200) | 2 | €81 | +€60 missed margin | €141 |
| Wednesday 14:00, positioning to Ruhr, 70 % backload prob, 350 km @ €1.25 | 3 | €81 | −€61 expected upside | €20 |
Same 100 km, same truck, same fuel, same driver. Three real prices: €81, €141, €20. Anyone treating them as the same number is going to make different decisions than the math would suggest.
What this changes operationally
Not "track empty-km percentage as a KPI." Every dispatcher with a pulse already does that, and the single percentage does not separate the three classes. The harder and more useful split is:
Look at the last 90 days of empty legs and classify each one into the three buckets. Roughly. The classification does not need to be precise; the ratios are what matter.
- If 60–70 % of empty km are Class 1, the fleet is well-tuned. Empty km exist because the network has unavoidable imbalances and the driver-hour planning is honest.
- If Class 2 creeps above 25 %, you have a dispatching problem, not a cost problem. The trucks are in the wrong places at the wrong times relative to where loads are appearing.
- If Class 3 is at 0 %, the fleet is not positioning strategically at all. That is fine for three trucks. For fifteen it is a competitive opening you are giving away to whoever in your corridor does the math.
Doing this classification on a quarter of empty legs tends to reveal a different problem than the one the fleet thought it had. A spreadsheet line showing 22 % empty kilometers reads the same whether that 22 % is mostly Class 1 (healthy) or mostly Class 2 (broken dispatch). The classification changes what you fix.
What this means for a single offer on the freight board
The right comparison is not "rate vs averaged CPKM" or even "rate vs loaded CPKM + positioning CPKM." It is "rate vs loaded CPKM + class-adjusted positioning cost."
- For a sunk-time empty leg, evaluate the offer aggressively. The empty is nearly free; the rate only needs to clear loaded CPKM by a small margin.
- For a peak-time empty leg, evaluate strictly. The empty is borrowing from a real alternative; the rate needs to clear both the direct cost and the alternative you are giving up.
- For a positioning-investment leg, evaluate the offer as the first leg of a two-leg portfolio, not in isolation. The right question is whether the combined margin of this load plus the expected backload clears both legs.
Freight-board offers do not arrive with the context to do this in your head. Where the truck is, where it is going next anyway, what alternatives exist nearby, what the backload market in the destination region looks like in the next 24 hours — these are inputs the offer itself does not give you, and the rate alone cannot tell you which class the empty leg belongs to.
Where tooling actually helps (and where it does not)
The honest version of this section: software cannot solve Class 3. It depends on regional market intelligence that varies week to week and is not cleanly available anywhere. What software can do is narrower but more important: make sure the direct cost number that feeds into the three-class frame is actually correct.
Most cost-per-kilometer figures in carrier spreadsheets are averaged or approximate. The €1.00/km in the worked example above only does its job if you actually know what those 100 km of positioning cost you, with the right fuel price for each country you drive through, the toll class matching your tractor-trailer combination, any ferry or tunnel segments, and the depreciation per kilometer on the specific vehicle assigned. Get any of those wrong and the Class 2 evaluation tips into the wrong decision more often than the framework saves you.
Transport Nomad's job in this picture is calculation precision. For any route you plan, the planner computes the direct cost of each segment (positioning and loaded) to the cent: actual fuel price per country, the correct toll for your axle count and emission class, ferries and tunnels at current operator rates. The three-class interpretation stays with the dispatcher. The cost number that feeds it stops being an estimate.
What no planner does honestly, ours included: predict whether the backload you are positioning for will actually materialize in 18 hours at the rate you expect. That part is still a judgment call.
What is still hard
None of this dissolves the genuinely difficult part of empty-km management: the call between accepting a marginal load now and holding empty for a better one in eight hours. That call depends on information you do not have at the moment of decision (the next load, the next rate, the next driver state), and any model that claims otherwise is selling something.
The three-class frame is not a formula. It is a way to stop reasoning about empty kilometers as if they were one thing. They are not, and the gut already knew it. The CPKM in the spreadsheet is the part that has not caught up yet.
If you want to see how this looks with your own corridors, you can try Transport Nomad free for 14 days. Or take the three-class frame and run it across the last quarter's empty legs in the spreadsheet you already have. The classification exercise alone tends to be worth a meeting.