When the Other Side Writes the Ticket

Pull a month of invoices and sort the tickets behind them into two piles: the ones your operation printed, and the ones somebody else did. For a lot of haulers, farms, and materials producers, the second pile is bigger than the owner expects. Gate tallies a contractor's grade checker signed. Destination weights off a customer's scale. Grain tickets the elevator printed, moisture and dockage already deducted. Broker load sheets. Each of those is a billable event documented by the party being billed, or by their agent, and that arrangement carries a cost that never appears as a line item anywhere.

"One ticket, one billable event" has a clause hiding inside it: somebody has to write the ticket. Whoever does controls three things at once: the count, because a load their record missed didn't happen; the quantity, because their instrument took the measurement; and the timing, because the record exists when they say it does. When the paper is yours, a dispute starts from your number. When it's theirs, there's usually no dispute at all, because you have nothing to dispute with.

The misses run one way

None of this requires anyone to cheat. The contractor's grade checker isn't keeping your books; he's keeping his. His tally exists so his project hits its yardage, and your getting paid is a side effect. When he's also directing trucks, answering a radio, and arguing with a surveyor, the tally is the task that slips.

Now notice the shape of the errors that produces. A counterparty's record can miss your loads. It almost never invents extra ones, because nobody signs for a load that didn't arrive. So the gaps in someone else's series don't scatter in both directions the way your own keying errors do. They lean, and they lean against you.

Put a number on it. A hauling outfit runs three trucks on a six-week structural fill job at $145 a load, paid off the contractor's gate tally. Month end comes. The drivers' own count says 214 loads. The signed tally says 206. Nobody can reconstruct the difference, because the only record that governs is the one with eight loads missing from it, and driver memory is not paper. You invoice 206. That's $1,160 gone, call it 4% of the month, and no report will ever surface it, because the governing record and the invoice agree.

A record with no force is still a record

The obvious fix is to write your own ticket for every load, and the obvious objection arrives right behind it: the contract says their number governs, so your ticket changes nothing and the paperwork is wasted motion. Plenty of dispatchers believe that. It's wrong, and the reason it's wrong is the useful part.

Your ticket can't override their tally. It can count against it. A driver who says "I think we ran more loads than that" is offering a feeling. A stack of 214 tickets, each with a date, a time, a truck number, and a gate, is offering eight specific holes: these dates, these hours, this truck. A grade checker shown eight timestamped gaps while the job is still open will mostly go look, and mostly sign. The same eight loads raised at final settlement, against a closed job and a demobilized crew, mostly get dropped.

Grain runs the same structure. The elevator's scale sets the weight, the elevator's probe sets the moisture, and both print on the elevator's ticket, with shrink and dockage taken before settlement. A probe that reads half a point wet doesn't announce itself; at a typical shrink schedule, half a point across 40,000 bushels of corn is roughly 280 bushels that came off your settlement, about $1,260 at $4.50. Nobody rigged anything. Instruments lean. A lean in the buyer's instrument is invisible to a seller who keeps no record of his own, and merely annoying to one who wrote his meter's reading on his own load ticket, load by load, and can set that series next to the elevator's and watch the difference hold steady for three weeks.

That's the general rule: a governing record can only be audited by a non-governing one. The second series is cheap to keep while the loads are running and impossible to reconstruct after they've stopped.

Make the comparison a habit, not a project

The second series earns nothing sitting in a folder. The value is in the meeting of the two, load by load, while the job is open, and that's a counting problem, not a judgment problem.

flowchart TD
  A["Load leaves your control"] --> B["Their record: tally, scale, probe"]
  A --> C["Your record: ticket at handoff"]
  B --> D["Their number becomes the invoice basis"]
  C --> E["Match the two series, load by load"]
  D --> E
  E --> F["Match: bill it and move on"]
  E --> G["Gap: chase it this week, while the job is open"]

Digitize both series, yours and theirs, and the checks are mechanical. A ticket in your series with no line in theirs gets flagged. So does a destination weight sitting more than a point and a half under your loaded weight, or a moisture number on their ticket that runs wet against the one on yours for the third straight week. A billing rule can say two things at once: their number governs, and a variance past this threshold surfaces for a person to look at. Reference data does the unglamorous part, mapping their site names, scale IDs, and ticket formats back to your customers and jobs so the two series can meet at all. That's the shape of what Termco does with this paper: digitize the tickets from both sides, flag the ones that don't hold up against your rules and the opposing series, and assemble invoice totals from lines that survived the check instead of lines nobody looked at.

There's a one-question audit hiding in all this. List every counterparty whose paper becomes your invoice basis, and for each one, ask when your own numbers were last compared against theirs, even once. If the honest answer is "never," that has a plain meaning: your real invoice basis is whatever they wrote down.

If you want a starting point, send 10-20 sample tickets to hi@termco.ai, including the counterparty's paper if that's what you bill from, and we'll send back what we find within 48 hours.