Fourteen Loads Past the Hold

On Tuesday the 6th you finally made the call you'd been putting off. Boyd Contracting, $11,480 past sixty days, promises since August: on hold. No more material until a check clears. Your office manager flagged the account in the accounting system before lunch, and everybody in the office felt better.

Boyd's trucks loaded fourteen more times. The last one crossed your scale on the 19th, thirteen days after your company had, on paper, stopped doing business with him.

The hold travels by paper

The flag your office manager set changed a field in a system only the office looks at. The scale window runs on a different document: a printed list of cash-only and no-charge accounts, taped up next to the tare weights, reprinted whenever somebody remembers. Yours still says September at the top. If you run more than one gate, multiply the list by the gates and ask which copy is oldest.

And even a fresh list has to get past the operator. Her job is throughput, forty trucks a morning, and Boyd's driver is a regular who hauls out of there three times a week. A hold on a stranger enforces itself; no account, no load. A hold on a regular is a favor she has to refuse in person, with a line idling behind him, on the strength of a list nobody told her changed.

Nobody failed here. The office did its job and the window did its job. The two jobs just read different paperwork, and nothing in the building carries a status change from one to the other faster than a reprint.

The decision was late before it was slow

The harder half belongs to the office outright: the 6th was already late.

Your aging report counts invoiced receivables and nothing else. On the morning you made the call, the keying tray held nine days of Boyd's tickets, loads hauled and signed for but not yet priced, call it $6,300 that existed only as carbon paper. His real tab wasn't $11,480. It was close to $17,800, and if your private trigger point was $12,000, he'd crossed it back in September. The report couldn't say so until the tray became invoices.

So the lag runs in both directions at once. Work flows into the office slower than it happens, and decisions flow out to the gate slower than they're made. Add up what the two lags did to one account: $6,300 unkeyed when you decided, then fourteen loads at around $420 each after you decided, another $5,900. By the time the window actually stopped Boyd, the balance you meant to cap at $11,480 stood near $23,700.

Every one of those tickets will key clean. Right customer, right product, right rate, invoiced on schedule. The paperwork is perfect, and what it documents is roughly $12,000 of credit that nobody approved, extended one ordinary load at a time.

flowchart TD
    A["Boyd's tickets, written daily"] --> B{"When does the office see them?"}
    B -- "keyed at month end" --> C["Aging says $11,480, true tab $17,800"]
    C --> D["Hold decided Oct 6, travels by printed list"]
    D --> E["14 more loads before the gate knows"]
    B -- "digitized the day they're written" --> F["Yesterday's tickets priced and totaled"]
    F --> G["Ticket 15 flags the morning it arrives"]

A limit nobody can look up is a wish

Boyd's credit application, the one in the folder with his references, says $15,000. Here's the claim worth arguing about: that number was never a limit. A limit only exists if somebody can check the balance at the moment it matters, and nobody in your operation can. The office sees a number that's nine days stale, and the window sees a list that's a month stale. The policy you actually ran was this: a customer may owe any amount, provided the paper lags.

Plenty of owners will push back, and the pushback is respectable. He can buy from two other pits, a hold costs the relationship, and loading a slow payer beats losing a steady one. Often true. But that's a decision to make with your eyes open, account by account, not an outcome the keying tray picks for you. Wanting a limit and running a wish look identical right up to the write-off, where the extra $12,000 gets filed under bad luck instead of under thirteen days of lag.

There's a quieter cost too. Boyd's bookkeeper now knows exactly how firm your holds are, because she watched this one not bite for two weeks. The next hold you place, on anyone, starts from that reputation.

Make the tray part of the number

Both lags close the same way: the tickets have to become data the day they're written, and account status has to live where pricing can read it.

That's the slice Termco builds. Tickets get digitized daily, as your drivers and operators already write them, so yesterday's loads are priced and totaled instead of sitting in a tray. Account status sits in reference data with a date on it: open, hold, cash only. A rule reads every incoming ticket against it, and a load on a held account gets flagged that morning as the exception it is, instead of surfacing at month end as six more invoice lines. Invoice totals assemble from the tickets that cleared. Nothing changes at the window: no terminal, no lookup screen between your operator and forty trucks, the same book and the same pen. The check runs behind the office door, which is where it belonged all along.

That flag isn't a gate arm. Boyd's fifteenth load would still have crossed the scale. It would also have been on your desk by the next morning, a named exception thirteen days sooner, while it was one load instead of fourteen.

Price your last three holds

The audit takes an evening. Pull the last three accounts you put on hold. For each one, write down the date the decision was made, then count two piles of tickets: the ones sitting unkeyed in the tray that day, and the ones dated after it. Price both piles. The total is the gap between the credit policy you set and the one your paperwork actually ran, and it repeats on every hold you'll place this winter.

If you'd like a second reader on the paper side, send 10-20 sample tickets to hi@termco.ai and findings come back within 48 hours.

Boyd might still pay, and if he does, the fourteen loads were revenue and this was a fire drill. Maybe. The thing to notice is that you didn't decide that. You decided to stop at $11,480, the paper stopped at $23,700, and everything between those two numbers was approved by nobody.