Nobody Audits the Freight Invoice

Ask a mid-market operations lead how much they spend on freight and you get a confident number. Ask how much of it was billed correctly and the room goes quiet.

The invoice arrives. It matches the rate agreement closely enough. Accounts payable has a run on Thursday. Someone signs it. Multiply that by four hundred invoices a month across six carriers, and you have a category of spend that is large, recurring, and effectively unexamined.

Why the audit never happens

Checking one invoice properly means reconciling the billed rate against the contracted tariff, verifying accessorials against what actually happened at the dock, confirming dimensional weight, checking fuel surcharge against the published index for the correct week, and matching all of it to the BOL and the proof of delivery.

Eight minutes when nothing is wrong. Forty when something is. Nobody on a lean team has eight minutes four hundred times a month, so the work gets triaged: check the large invoices, spot-check the rest.

That is a reasonable adaptation to constrained attention, and precisely the shape of process that leaks. Billing errors are not concentrated in the large invoices. Duplicate accessorials, misapplied residential surcharges, reweighs never reversed, detention billed on a window you did not cause — small, unremarkable, relentless.

What it actually costs

Direct overpayment. Small per instance, material per year.

Lost recovery windows. Carrier contracts give you 30 to 180 days to dispute. An error found in month seven is not an error, it is a donation.

Blind negotiation. The expensive one nobody counts. At renewal you negotiate from your own invoice data. If that data carries systematic errors you never caught, you are calibrating against a distorted baseline. A rate concession on a lane where you have been quietly overpaying accessorials for two years is not a win.

Why hiring does not fix it

A credentialed logistics professional does not want to spend their week checking fuel surcharge tables — hire them for that and they leave. Hire someone who will happily do it and they miss the contractual subtleties, which is where the money is. Either way the volume consumes all the capacity, leaving no time for the work only they can do: challenging a pattern of misclassification, restructuring an accessorial schedule, deciding whether a dispute is worth the relationship friction.

And volume is spiky. Peak season does not care about your headcount plan.

Split throughput from judgement

The throughput half is systematic — every invoice, every line, every time: rate against tariff, accessorial against event record, weight against manifest, surcharge against the correct index period, invoice against BOL and POD. Flag, categorise, assemble the dispute package. High-volume, rule-governed, unforgiving of inattention: the combination humans handle worst. Full coverage, before payment, not a sample after it.

The judgement half is not systematic. Whether a variance is an error or a charge you did not anticipate. Whether a recurring misclassification is a carrier defect or your own dock coding shipments wrong. Whether a pattern justifies escalation or renegotiation. Those calls belong to someone who has sat across the table from a carrier.

Under the h.work model the AI Specialist runs the audit at full coverage; the logistics expert reviews exceptions and owns the consequential decisions. Every correction becomes part of what the Specialist knows, so the exception queue shrinks instead of holding steady. Manager tier, $3K–$6K per month, against a $90K–$160K loaded hire. Monthly terms, first month a paid trial, deployed into your TMS and accounting stack.

The test

Pull twenty carrier invoices at random — not the big ones. Audit them properly. Two numbers matter: what percentage contained a billable error, and what percentage of those are still inside the dispute window.

If the first number is not close to zero, you have a throughput problem, not a diligence problem. And throughput problems do not get solved by asking a busy team to try harder.