Every air cargo quote contains a forecast. The shipper states a weight and a set of dimensions, the forwarder computes a chargeable weight, and a price is attached to it. Weeks later the freight arrives at the warehouse, goes over a scale and under a dimensioner, and a different chargeable weight comes out. The industry treats the second number as the truth and the first as paperwork. It should treat the first as a forecast and the second as the score.
Chargeable weight is the greater of actual weight and volumetric weight, where volumetric weight is volume in cubic centimeters divided by 6,000, or about 167 kilograms per cubic meter. Most general cargo sits near the line.
The error is not random
If declaration errors were noise, they would cancel over a season and nobody would care. They do not cancel, for 3 reasons that all push the same way.
Shippers estimate dimensions from the pallet footprint and the nominal stack height. They do not add overhang, stretch wrap, corner boards, or the gap where the last layer did not fill. They round down, because the person filling in the booking has no reason to round up. And the incentive runs 1 direction: declaring less never costs the shipper anything at booking.
Take an illustrative shipment: 1,000 kilograms on 4 pallets declared at 1.2 by 1.0 by 1.25 meters each, 6 cubic meters, about 1,000 kilograms volumetric. Chargeable weight at quote, 1,000 kilograms. At tender the dimensioner reads 1.25 by 1.05 by 1.32 meters on each pallet, 6.93 cubic meters, 1,155 kilograms volumetric. Chargeable weight, 1,155 kilograms. The shipment grew 15.5 percent between the quote and the aircraft, and the growth came from 5 centimeters here and 7 centimeters there.
Two desks pay for the same error
The pricing desk pays first. If the forwarder bills on measured weight, the revenue arrives, but the margin was computed on 1,000 and the capacity was bought on 1,000. The extra 155 kilograms are argued over, invoiced late, or waived to keep the account.
The capacity desk pays second, and pays more. A block of 40 tons a week planned at 70 percent utilization leaves 12 tons of headroom. If booked freight runs 8 percent over declared volume on average, the plan has quietly consumed more than 2 tons of that headroom before the week begins. The freighter bulks out before it weighs out, cargo is offloaded, and the offload is charged to bad luck rather than to a bias that was visible in 6 months of records.
Forecast it like a forecast
Once the gap is named as a forecasting problem, the tools are standard. Measure the bias per customer, per commodity, per origin: the ratio of tendered chargeable weight to quoted chargeable weight over the last 50 shipments. A customer at 1.06 over 50 shipments is not noise. The quote for that customer should be priced, and the space reserved, at 1.06.
Density matters too. A commodity whose density sits within 10 percent of 167 kilograms per cubic meter will flip between weight charging and volume charging on small measurement changes, so its forecast error is largest and its quote should carry the widest margin or a measured weight clause.
Then close the loop. The dimensioner at tender is where the forecast is scored, and the score belongs in the quote system, not in the warehouse system. In the operations we have seen, the measurement exists, the quote exists, and the join between them has never been built. The bias is sitting in 2 tables that have never met.
A quote is a price attached to a forecast of what will arrive. Treating the arrival as the truth and the quote as paperwork gets the roles backward. The quote is the decision. The scale is the feedback.