Billed Off the Board

It's the 28th, the Route 9 hauling invoice is due, and your billing clerk has two documents to build it from.

One is the ticket stack: 50-some carbon tickets in a folder, a dozen more still riding in doorpockets, two with diesel thumbprints where the tonnage goes. The other is the dispatch sheet. Clean columns, one line per load, already totaled by truck and by day, sitting in a binder eight feet from her desk.

She bills the sheet. Of course she does. The invoice that goes out is a careful, legible record of what was supposed to happen three weeks ago.

The board is a plan, not a record

At 6am the dispatch board is honest. Nine loads of base to Route 9, three trucks, start at the pit. Then the day goes to work on it.

At 1:40 the paving foreman gets ahead of schedule and radios for two more loads before the plant closes. Dispatch shuffles a truck off another job and sends it. At 2:15 the paver throws a track, and the last two scheduled loads die in the pit with rock in the bed, dumped back on the pile.

None of that goes back on the morning's sheet, because the sheet already did its job. Its job was to start the day. Dispatchers fix tomorrow's plan, not yesterday's paperwork, and the radio call that added two loads took nine seconds, which is why the radio got used. So by Friday the board says 54 loads. The trucks also ran 54 that earn money, and they aren't the same 54: two were born on the radio and never touched the board, and two the board swears by died in the pit. The totals reconcile to the load. Four lines are wrong, and the only documents that know which four are riding around in doorpockets.

Tickets are written by the event itself. A load that runs gets a ticket whether or not any board expected it, and a load that dies gets none, however firmly it was scheduled. That's the whole argument for building invoices out of them: one ticket, one billable event, in both directions.

It leaks both ways at once

Bill the board and you leak twice, in opposite directions, on the same invoice.

The add-on loads vanish. Two radioed loads a week, 23 tons each at a $9.25 haul rate, is about $425 a week on one job. Call it $5,500 over a season if the pattern holds, and it usually does, because hot afternoons are when foremen radio for more. The work got done and the driver got paid. The revenue was never anywhere a month-end clerk could see it, because the only record of the order was a nine-second radio call and the only record of the haul is a ticket nobody opened.

The cancelled loads get billed. That one's worse than its face value. The GC's project engineer matches your invoice against his gate log, finds two loads his job never received, and now you're issuing a credit memo and he's auditing every invoice you send for the rest of the contract. An underbilled customer stays quiet. An overbilled one builds a process around you. Billing from the plan manufactures both kinds, every month, from the same binder.

flowchart TD
  A["Dispatch board: 54 loads planned"] --> B{"What builds the invoice?"}
  C["Tickets: 54 loads ran, not the same 54"] --> B
  B -- "the board" --> D["2 radioed loads never billed"]
  B -- "the board" --> E["2 dead loads billed, credit memo follows"]
  B -- "the tickets" --> F["Invoice matches what the trucks did"]

Why the board keeps winning

Nobody ever decided the board would be the billing record; it won on availability. On the 28th the board is complete, totaled, and in one place. The tickets are still trickling in, half of them written at arm's length on a steering wheel. A clerk with an invoice deadline doesn't pick the better document; she picks the one that's there. Paper doesn't lose to software in these offices. It loses to deadlines.

Your dispatcher will defend the sheet, and his best argument deserves a straight answer: "my board is more accurate than what my drivers hand in, because drivers lose tickets." Load for load, he might be right. He's still wrong about the job it can do. The board is kept by the same person who made the plan, so it grades its own homework. It can't record a load nobody told dispatch about, and plenty of loads are born at the job site, not the office. And when the GC disputes a line, a board entry is your word in your handwriting. A ticket signed at the gate is a record both companies already accepted.

Here's the version a skeptic will actually argue with: a disciplined operation should bill nothing off the dispatch board, ever, including the weeks the ticket stack comes back ragged. A ragged stack is a problem to fix at the stack: chase the doorpockets Monday, not the 28th. Promoting the board to a billing record doesn't fix missing paper. It just makes missing paper invisible.

Let the two documents argue

The 2010s answer was to put the board in the cab: an e-ticketing app, every driver logging every load on a phone, so dispatch and billing share one screen. Drivers didn't want it, and haul roads don't carry signal. Ask an outfit that ran one where the app tickets went at month end; the honest answer tends to be a third record to reconcile against the first two. The carbon ticket and the radio were never the weak links. Keep them. The weakness is that the board and the tickets never get compared while anyone still remembers the week.

So compare them while they do. Tickets digitized the day they come in off the trucks, as written, thumbprints and all. Haul rates and zones keyed as billing rules, jobs and accounts as reference data, so each ticket prices itself, and a ticket that breaks a rule, say a load with no open job behind it, gets flagged that day instead of surfacing on the 28th. Invoice totals assemble from the tickets that cleared. That slice between the doorpocket and the invoice is the part Termco does, and none of it reaches the cab, the pit, or the board. What it leaves on your desk is the half of the argument you never had on time: a priced ticket count by Tuesday. Lay that against the board while dispatch can still say this one was the paver track and that one was the foreman's radio call, and the board goes back to being what it was always good at, which is tomorrow morning.

Two counts, ten minutes

Run the test on last month. Take your biggest hauling job, count the lines on the invoice, then count that job's tickets in the folder. If the second number is bigger, you underbilled; if it's smaller, find out which loads you invoiced on faith. Then ask your dispatcher how many loads got added by radio last week, and go find each one on paper.

If you want a second set of eyes on the stack, send 10-20 sample tickets from one job to hi@termco.ai and findings come back within 48 hours.

The board earned its place on the wall. It starts the day better than any document you own. It just shouldn't be allowed anywhere near the 28th.