Assigning Loads to Drivers: The Margin Math Nobody Runs

Almost every article about picking loads was written for a soft market, where the job is finding freight at all. That is not the market you are dispatching in right now. DAT's Dry Van Report published August 3, 2026 put the dry van load-to-truck ratio at 10.93 for the week ending July 31, 2026, up from 9.92 the prior week and 6.64 a year earlier. Nearly a two-thirds jump year over year.

When there are almost eleven loads for every truck, you are turning freight down all day whether or not you are doing it on purpose. Dispatch stops being a scheduling job and becomes an allocation job. Most small carriers have no repeatable way to make that call, so they default to the highest rate per mile on the board, which is usually the wrong number.

How should a dispatcher decide which driver gets which load?

Assign on total contribution across the whole cycle, not on the posted rate per mile. That means loaded revenue minus the cost of getting to the pickup and the cost of getting out of the delivery, divided by the hours the load will actually consume for that specific driver.

The trap is simple to state and hard to resist: the highest rate per mile on the board is frequently the lowest-earning load of the day.

That matters more now than it did last summer. In the same week that the ratio hit 10.93, DAT reported dry van spot linehaul averaging $2.32 per mile, down 2.4% or six cents from the prior week, and noted those rates are linehaul only, excluding fuel and surcharges. More choices, not automatically better ones. That combination is exactly when selection discipline pays.

Why does the posted rate per mile mislead?

Four things the posted number leaves out, all of which come out of the same margin.

It counts only loaded miles. The deadhead to the pickup is real fuel, real wear, and real hours on a driver's clock, and it appears nowhere in the rate you are looking at. A $2.60 load with 180 miles of deadhead in front of it is a different load than a $2.40 sitting under the truck.

It ignores where the truck ends up. A strong rate into a lane with nothing coming out means the next leg is cheap, or empty. You paid for the good rate with the bad one behind it.

It ignores time. Two loads at the same rate are not the same load if one goes to a receiver with a four-hour dock reputation and the other does not. Hours the driver spends at a dock are hours that produce nothing.

It ignores when the money arrives. A carrier funding fuel this week is running a different business than one with a 45-day cushion. Payment terms are part of the rate, whether or not the load board treats them that way.

What four numbers actually decide the load?

Strip it down and there are four inputs. None of them require a spreadsheet.

• All-in revenue for the move. Include the accessorials you can realistically document and collect, not the ones you hope for. Detention you cannot prove is not revenue.
• Total miles, including deadhead in and likely repositioning out. The honest denominator, not the flattering one.
• Hours the load consumes for that specific driver, given where they sit right now and what they have left. This is a planning judgment the dispatcher makes. Howdy Dispatch is not an ELD and does not calculate hours of service.
• Days to pay for that broker. You already know this from your own history with them, even if it lives in someone's head rather than a report.

Here is an illustrative comparison, not a real carrier's numbers. Load A pays $2.65 per mile over 400 loaded miles, which is $1,060, but it sits 150 deadhead miles away and delivers into a region where your last three trucks waited two days for a decent reload. Load B pays $2.35 over 520 loaded miles, which is $1,222, with 20 miles of deadhead and a delivery 40 miles from a market you can always cover out of.

Load A has the better rate per mile and loses. Once you count the deadhead, Load A runs 550 total miles for $1,060, or $1.93 all-in. Load B runs 540 for $1,222, or $2.26. And Load B leaves the truck somewhere useful. The board would have shown you A first.

How does driver fit change the answer?

The same load is a different load depending on who runs it. Where the driver is sitting right now, how close it puts them to home, what they have already run this week, what equipment is on the truck.

This is the part that spreadsheets miss and experienced dispatchers know in their gut. A load that pays well and sends a driver the wrong direction on the wrong week is how a small fleet loses a driver it spent three months finding and six weeks training. That cost never shows up on the load, but it is real and it is larger than the margin you captured.

Say it plainly: the dispatcher makes this call. Software's job is to put the inputs in front of you fast enough that you can make it well. It does not decide, and any vendor telling you their system picks loads for you is selling you something that will eventually pick a bad one. We do not do route optimization, autonomous dispatching, or broker matching, and we are not planning to pretend otherwise.

What a small fleet needs on the board to make this call in two minutes

The decision above is not hard. What makes it hard is that the inputs are scattered across a phone, a text thread, a PDF nobody opened, and one person's memory. To make this call in the two minutes you actually have before the load is gone, four things have to be in one place.

• Where every truck is right now, without making a phone call to find out.
• What each driver is already committed to, in one view rather than in a text thread you have to scroll.
• The rate confirmation's terms already read and attached to the load, so you are comparing terms instead of opening PDFs. This is where our AI dispatch platform does real work: you attach the broker's rate confirmation and it pulls out the customer, origin, destination, mileage, rate, pickup and delivery times, and broker contact, then pre-fills the load for you to review. Five to ten minutes of typing becomes about twenty seconds of checking. That is the only AI feature we claim as live today, and it is deliberate.
• Your history with that broker on the last few loads, including whether the paperwork came back clean.

None of that decides the load. All of it is why you can decide the load before someone else books it.

The habit that separates the fleets that grow

Decide the rule before the phone rings.

A carrier that knows its floor per hour, or per all-in mile, says no faster and books better freight. The carrier without a floor negotiates every load from scratch, which takes longer and lands lower, because the broker on the other end absolutely has a number and you do not.

Write the floor down and revisit it when the market moves. A floor set during a soft stretch is the wrong floor in a 10-to-1 market, and a floor set now will be wrong again when capacity comes back. The number is not permanent, the habit of having one is.

Then track what you declined, not just what you booked. Your declined column is the only place you find out whether your floor is right. If you are declining nothing, the floor is too low. If you are declining everything and the trucks are sitting, it is too high. Booked loads alone will never tell you that, because they are the ones that cleared the bar by definition.

A load board tells you what freight exists. It does not tell you which of those loads is right for the truck you have free on Thursday, with the driver who needs to be home Saturday. That gap is the whole job, and it is worth more than a tenth of a cent per mile.

Want the board that makes this call fast? Start a 14-day trial, no card required.