Empty Miles Per Loaded Mile: What Deadhead Costs Now
Empty miles per loaded mile is the number most small fleets never run. With rates and costs both at record levels, here is the math and what it changes.

Rates are the best they have been in years and plenty of small carriers still are not making money. Deadhead is where a lot of that difference hides.
What is empty miles per loaded mile, and why does it matter now?
Empty miles per loaded mile is exactly what it sounds like: deadhead miles divided by loaded miles. It is the ratio that turns the rate you quoted per loaded mile into what the truck actually earned for every mile it turned.
The reason it matters more this year than last is that the spread got thin at both ends. ACT Research reported in late August that dry van, reefer, and flatbed spot rates excluding fuel ran 41 to 47 percent above year-ago levels in July 2026, with aggregate contract rates up 17 percent and contract rates excluding fuel at $2.50 per mile after an eight-cent gain during the month. Good news, and it is why the market feels better than it did.
Now the other end. ATRI's operational costs report, released in July 2026 on 2025 data, put the industry average cost to run a truck at a record $2.336 per mile, up 3.4 percent, or $1.854 per mile excluding fuel, up 4.2 percent. Truckload and refrigerated operating margins improved slightly and still came in under 1.0 percent.
Read those two together. A contract rate around $2.50 against an average cost around $2.34 leaves roughly sixteen cents per mile, and that is before your deadhead is counted. When the margin is that thin, the denominator decides whether the week works.
The math, run on a real load
Say Copper Ridge Logistics tenders you a run from Dallas to Memphis. Four hundred fifty loaded miles, $1,150 on the rate confirmation. Your dispatcher does the arithmetic every dispatcher does:
$1,150 divided by 450 loaded miles is $2.56 per loaded mile.
That is a fine number. Above contract average, comfortably above the cost figure, and it looks like a load worth taking.
Except the truck delivers in Memphis and the next pickup is in Little Rock, 137 miles west. Nobody pays for those 137 miles. The truck still burns fuel across them, still wears tires, still runs the driver's clock.
So run it again on total miles:
$1,150 divided by 587 total miles is $1.96 per total mile.
Against an industry average cost of $2.336 per mile, that load lost about $221. The rate confirmation said $2.56. The truck experienced $1.96. The gap between those two numbers is the whole point, and it is 30.4 percent deadhead.
A caution worth stating plainly: $2.336 is ATRI's industry average on 2025 data, not your cost. Your fuel, your insurance, your equipment age, and your driver pay make your number different, possibly a lot different. Use the industry figure as a yardstick while you build your own, and build your own from your P&L, not from an article.
Here is the part that surprises people. Take that same load and cut the reposition to 45 miles, a 10 percent ratio. Now it is $2.32 per total mile, and at industry-average cost it is roughly break-even. A load paying $2.56 per loaded mile is a break-even load at 10 percent deadhead and a losing load at 30 percent. Nothing about the freight changed. Only the empty leg did.
Run it the other direction if you want the number that actually protects you. To clear $2.336 all-in at a 30 percent deadhead ratio, that load needed about $3.05 per loaded mile, not $2.56. That is the number to carry into the call.
Why dispatchers rarely run this number
It is not ignorance. It is that every artifact in the workflow is written in loaded miles.
The rate confirmation states loaded miles, because that is what the broker is paying for. The load board posts loaded miles. Your invoice goes out on loaded miles. When the entire paper trail speaks one language, that language becomes the way you think about the truck.
The empty leg also happens at the worst possible moment for attention. It occurs after the load is closed, when the job is done, the POD is in, and everyone has moved on to the next tender. Nobody is looking at a load that is already delivered.
And in a spreadsheet, the empty leg is simply not a row. It has no rate, no broker, no load number, no confirmation. There is nothing to type. A row that never gets typed is a cost that never gets counted, which is how a fleet can run 25 percent deadhead for a year without a single number on a page saying so.
How do you actually measure deadhead in a small fleet?
You already have the inputs. It is a records problem, not a math problem.
For any two consecutive loads on the same truck, you need four things: where and when the first load delivered, and where and when the next one picked up. The miles between those two points are your deadhead. That is the entire calculation.
The obstacle is that reconstructing those four data points from memory, texts, and a folder of photos three weeks later is miserable, so nobody does it twice.
This is where having the timeline already recorded changes what is realistic to measure. Howdy Dispatch keeps continuous GPS tracking on the HQ live map, and drivers close out pickup and delivery with one tap and a photo, so every load carries its own timestamped record with the documents attached. The gap between the last drop and the next pickup is visible in the record instead of reconstructed from somebody's recollection.
Be clear about the boundary, because this matters. Howdy Dispatch is not a route optimizer and it does not decide which load to take. It does not compute your deadhead ratio for you and it will not tell you to reject a tender. It keeps the timeline and the per-load record straight so that you, the person who knows your lanes, can run the math and make the call. We are also not an ELD, not a freight broker, and not a compliance product. This is a dispatch record, not a regulatory one.
One practical note on how to look at it: run a weekly rollup by truck, not a per-load autopsy. Any single week has a bad reposition in it, and chasing one ugly load teaches you nothing. A truck sitting at 12 percent and a truck sitting at 29 percent over a month is a real signal you can act on. Look at the pattern.
What changes when you start watching it
Three things tend to shift once the number is in front of you.
The best-paying load stops being obviously the best load. A tender at $2.90 per loaded mile that strands the truck 200 miles from anything can lose to a $2.35 tender that delivers into a market where you reload the same afternoon. You cannot see that on a rate confirmation. You can only see it once the empty leg has a number.
The reload decision moves earlier. The value of booking the next load before the trailer is empty is precisely the deadhead you avoid, and now you can put a dollar figure on it. That reframes the argument about whether it is worth spending twenty minutes on the phone Thursday afternoon instead of Friday morning. At $2.336 a mile, killing a 137-mile reposition is worth about $320.
Lane discipline starts beating rate chasing. Two decent lanes that touch each other will usually beat one great lane that strands the truck. This is the least intuitive of the three and the most durable, especially now. ACT noted in the same August report that the Class 8 tractor fleet has contracted and available equipment remains historically tight, which is a large part of why rates recovered. Tight capacity rewards the carrier who keeps a truck productive over the one who wins the highest single rate.
If you want to see the whole picture, HQ analytics and the per-load document access sit alongside the live map, and the driver side of that record comes from the iOS app the driver actually uses rather than an HQ dashboard the drivers ignore. For what the software does and exactly where AI does and does not show up, that is all laid out on the AI dispatch platform page.
Something to do Monday morning
You do not need us for this. Pull your last twenty completed loads. For each consecutive pair on the same truck, write down the delivery city and the next pickup city, and get the miles between them however you normally would. Add the empty miles, add the loaded miles, and divide.
Whatever number comes out is the one that has been quietly setting your margin all year. If it is in the teens, your dispatch is tighter than most. If it starts with a two or a three, you now know where this year's money went, and you know it before your accountant tells you in April.
Then take your worst load from that list, and recompute what it needed to pay per loaded mile to clear your cost all-in. That is the number to bring to the next call with that broker.
When you want the record to keep itself instead of rebuilding it from texts and camera rolls, start a 14-day trial and run a week of loads through it.
READ NEXT
Tender Rejection Rate: What 14% Means for a Small Carrier
Tender rejection rate hit 14.4% in September 2026. What a rejected tender is, why it lands on a small carrier's phone, and how to take it without losing money.
Diesel Price Spike: What It Does to a Small Fleet's Margin
Diesel near $5.60 a gallon changes which loads pay. The margin math behind a diesel price spike for a small carrier, and the paperwork that gets it paid.


