Cost Per Load: The Dispatch and Billing Hours You Spend on Every Single Load

The problem: You know your cost per mile down to the cent and have no idea what a load costs to run through the office, because those wages sit in overhead rather than against the freight.

The solution: Count the office steps that repeat on every load, price them, and put them where they belong, next to the revenue that load produced.

Every carrier this size can tell you cost per mile. Fuel, driver pay, tractor and trailer, insurance, maintenance, tolls. It is calculated carefully, it is watched, and it is the number the whole business is priced from.

Almost none of it includes the office. Dispatch wages, billing wages, and the hours spent on the phone about a load that already delivered land in overhead, get divided by total miles if they get allocated at all, and disappear into an average. So the cost of a load, as the business understands it, is the cost of moving the freight and not the cost of processing it.

That is a meaningful omission, because processing cost is not evenly spread. Some loads consume the office and some barely touch it, and there is currently no way to tell which is which.

The same steps, on every load, forever

Walk a single load from the call to the cash and the repetition is obvious once you look for it.

The load gets taken and entered. The rate confirmation gets sent, signed, and filed. A driver gets assigned and told. The shipper is called for an appointment. Somebody checks in on it at pickup. The customer wants a status, so somebody sends one, then sends another. Something goes late, so somebody explains it. It delivers. The paperwork comes back, eventually, in a photo, a scan, or a folder from a truck stop. Somebody reads the delivery receipt, matches it against the rate confirmation, checks the accessorials, and builds the invoice. Somebody files it in the customer's portal, which has its own format. Somebody follows up when it ages past 30 days.

Not one of those steps requires judgment on a routine load. Every one of them requires a person, because the information needed to do it lives in a different place than the person doing it.

The steps are known. What they cost is not.

Here is the specific thing a carrier at this size cannot answer: what did the office spend on load 88421.

The wages are known. The number of loads is known. What is missing is any connection between the two, which means all of the following stay invisible:

• Which customers are cheap to serve and which ones cost you an hour of somebody's day on every load, before you get to the rate.
• What a portal customer costs compared to one who accepts an emailed invoice, which is a real and often large difference.
• How much of the billing week goes to loads that were fine, as opposed to loads that need attention.
• Whether the next dispatcher is a growth hire or a symptom.

A customer paying a good rate per mile can be your worst account once you count the status calls, the reformatting, and the 60-day pay. That reversal is common and it is not visible in any report a carrier this size currently runs.

Why you cannot simply automate it as it stands

The obvious response is to hand the repeated steps to software. That is the right instinct and it fails on the first attempt for a reason worth understanding, because the reason is also the diagnosis.

Software cannot do these steps as they currently exist, because three things they depend on are not written down anywhere.

The first is information it can rely on. The load lives in the dispatch system, the rate lives in a signed PDF in an inbox, the driver's status lives in a text message, and the accessorial lives in something the driver said on the phone. There is no version of the load that is complete and current, so anything acting on it is acting on a fragment.

The second is the rules your people already follow. Your billing clerk knows what makes a load ready to invoice, and what to do when the delivery receipt is missing a signature, and which customer will reject a detention charge without a gate log attached. That knowledge is real and reliable and has never been said out loud, so it cannot be applied by anything except her.

The third is who is allowed to do what. Sending a status update is safe. Changing a rate is not. Until that line is drawn explicitly, the honest answer to whether software should be allowed to act is no.

None of those three are exotic. They are the reason the work is done by hand, and they are what you assemble first: the loads, customers, rates, and driver status connected into one current picture of the operation, the rules stated in plain language, and a clear line around what may happen without a person. Once those exist, the routine load can move through the office on its own, and the office spends its time on the exceptions, which is where the money actually is.

A look at a trucking company

Consider a carrier doing about $12 million a year, 45 trucks, mostly regional dry van with a handful of committed lanes. Three dispatchers, a billing clerk, and an office manager. The owner watches cost per mile weekly and has never calculated cost per load, because nobody does.

Suppose the four office people log their time by activity for two weeks. You would expect roughly an hour a day each to fall into the repeat category: status updates for customers who could have looked it up, retyping rate confirmation details into the dispatch system, rebuilding an invoice from paperwork that arrived as a photograph, and keying the same invoice into a customer portal in a second format. At a loaded rate of about $28 an hour, that is on the order of $28,000 a year of wages producing nothing a customer pays for.

The more valuable finding would be the distribution. Sort those hours by customer and you would expect the picture to be lopsided: a small number of accounts consuming a large share of the office. The customer with the strict portal, the manual detention approval process, and the weekly status call could plausibly cost several times the office minutes of a comparable account at the same rate per mile. Nothing in the current reporting would ever surface that, so that customer has been quietly repriced by nobody.

What an owner would likely do with this is not fire the billing clerk. It is take the two worst accounts into a rate conversation with an actual number in hand, stop adding office headcount every time volume steps up, and move a dispatcher's freed hour onto the loads that need a person, which is the hard reroute, the difficult shipper, and the driver who needs a call.

How to start

1. Log office time by activity for two weeks. Split repeated steps from exceptions, and note the customer on each.
2. Divide by loads, not by miles. Cost per load is the number that reveals this, and cost per mile is the number that hides it.
3. Rank your customers by office minutes. Then set that ranking next to your rate per mile ranking and see how much they disagree.
4. Mark the steps that are pure transfer. Anything where a person is moving information that already exists somewhere else is the removable set.
5. Fix the load record before automating the step. A routine load has to be complete and current in one place first, or you are only automating the retyping.

The takeaway

Your cost per mile is precise about the truck and silent about the office, so the wages buried in every load have never been counted or charged to the freight that caused them. Count them for two weeks, divide by loads rather than miles, and rank the result by customer. You are likely to find both a payroll number nobody in the business has seen and at least one good-rate customer who has been expensive all along.