Deadhead Miles: How AI Weighs Empty Miles in Load Decisions

Every empty mile is a mile you pay for and nobody pays you back. Fuel burns, the driver's clock runs, tires and brakes wear, and hours of service tick down — whether or not there's freight in the trailer. Yet most load decisions still start with the posted rate per mile, a number that says nothing about how far the truck has to run empty to earn it. That gap, between the posted rate and the all-in rate, is where margin quietly disappears.

What deadhead actually costs

Deadhead is any mile driven without a revenue load, and it comes in two forms. The first is the run to a pickup — positioning the truck where the next load starts. The second is repositioning: driving out of a weak market toward freight, or home for a reset, with nothing in the trailer to cover the move. The first is easy to price into a booking because both endpoints are known. The second is strategic, and sometimes worth it — 200 empty miles to reach a lane that pays for two weeks is an investment, not a loss.

Either way, the meter runs. ATRI's 2025 operational-cost research, on 2024 data, puts a truck's average marginal cost at about $2.26 a mile once you add up fuel, driver wages and benefits, maintenance, insurance, and equipment. At that number, 120 empty miles to reach a pickup costs roughly $270 before the load earns a dollar. Industry estimates generally put empty miles somewhere in the 15–30% range of total miles, so on a truck running 100,000 miles a year, that's tens of thousands of miles of pure cost that never show up on an invoice.

The problem isn't that deadhead exists; some of it is unavoidable. The problem is that it's invisible at the moment of the decision. A dispatcher scanning a board sees $2.10 per mile and books it, then learns the pickup is 120 miles away in the wrong direction. The rate looked strong; the trip wasn't. A high deadhead share is rarely random — it usually traces back to something fixable, like weak backhaul planning or booking on the posted rate without counting the miles to reach the load.

Why the posted rate lies

The number that actually matters is the all-in rate: total revenue divided by total miles, loaded and empty. Consider two loads offered to the same truck:

Load A posts eight percent higher and pays out worse. Fold in the empty miles and the "cheaper" load returns nearly twenty cents more per mile run. A dispatcher comparing posted rates picks A; a dispatcher comparing all-in rates picks B — and the truck earns more for the same day of work.

The math is simple: all-in rate equals total load revenue divided by loaded miles plus deadhead miles. On Load A, that's $1,050 ÷ 620 = $1.69 per mile. The discipline is running it on every load you weigh and ranking by that number instead of the posted one. Two refinements sharpen it — subtract your cost per mile to see margin rather than revenue, which puts Load A underwater against a $2.26 all-in cost, and estimate the deadhead after delivery, not just before pickup, because where a load leaves the truck sets up the next empty run. Doing this by hand for one load is easy. Doing it for every load across eight trucks, in the seconds before a good one is gone, is where it breaks down.

How Numeo prices deadhead into the decision

Numeo's AI Hub ranks loads on the all-in rate, not the posted one. It already knows where each truck comes free and when — from the delivery point and appointment on the active load, or the availability a dispatcher sets — so when a load appears on a connected board, it computes the deadhead to that pickup, folds it into an effective rate across total miles, and ranks what's left. The loads at the top are the ones that still pay after the empty miles come out. A strong posted rate 150 miles away can land below a modest rate 15 miles away, because the math says so, and the dispatcher sees that before dispatch instead of after.

Destination matters as much as origin. Some lanes are structurally imbalanced — plenty of freight in, almost nothing back out — and a load that pays well into a region with no outbound volume can cost more in the next deadhead than it earned. Numeo weighs destination demand so the truck lands where the next load is easy to find, not stranded chasing freight out of a dead zone. On an imbalanced lane, that's the difference between a profitable week and 300 empty miles spent recovering.

Where the dispatcher still decides

None of this removes judgment from the desk. The math ranks loads; it doesn't book them. A driver heading home for the weekend may take a deeper deadhead on purpose. A standing broker relationship can be worth more than twenty cents a mile. An appointment that looks clean on paper may be a detention risk a dispatcher already knows to avoid. Numeo's job is to make the empty-mile cost visible at the moment of the decision — the all-in rate, the deadhead, the destination market — and leave the call to the person weighing everything that doesn't fit in a table.

The takeaway

Deadhead never shows up on a rate confirmation, which is exactly why it erodes margin so reliably. The fix isn't to chase the highest posted rate or to fear every empty mile — it's to see the all-in rate before you commit, every time. Price deadhead into the decision instead of discovering it after, and the truck stops taking loads that only looked good on the board.

See how AI Hub ranks loads on all-in rate, or search every board at once with Load Hub.