Passenger aviation solved dynamic pricing decades ago. The seat next to you cost a different amount than yours, and nobody finds this strange. Air cargo, flying in the same aircraft, still transacts a large share of its volume against rate sheets that update quarterly, distributed as spreadsheets, discounted ad hoc by sales teams working from memory and relationship. Meanwhile the actual value of a kilogram of capacity on a given departure moves hourly with bookings, cancellations, weather, and what competitors are quoting.
The result is systematic yield leakage in both directions. Underpricing on constrained departures gives margin away to whoever books first. Overpricing on soft ones flies air. The leakage is largely invisible because the counterfactual, what the capacity could have earned under state-aware pricing, never appears in any report.
What live yield actually requires
Dynamic pricing in cargo is harder than in passenger, not easier. Shipments vary in weight, volume, and density, so the unit of capacity is two-dimensional and interacts across bookings. Demand arrives lumpy through multiple channels with different elasticities. Show-up is uncertain: booked cargo materializes at variable tonnage or not at all. A serious pricing engine therefore needs a probabilistic forecast of demand-to-come by segment, a probabilistic model of show-up, the remaining capacity state in both weight and volume, and an optimization layer that turns all of it into a bid price: the minimum acceptable rate for the next kilogram and cubic meter on this departure, right now.
From rate sheet to live yield.
Same demand, two pricing policies, thirty days to departure.
The rate sheet leaves money on the table twice: early when it overprices, late when it underprices.
The organizational half of the problem
The mathematics is necessary and insufficient. Rates flow through people: sales teams with delegated discretion, long-standing customer agreements, allocation commitments to forwarders. A pricing engine that ignores this gets ignored in return. The systems that stick present the bid price as a floor with explainable reasoning, preserve negotiated frameworks while pricing the spot dynamically, and show sales the opportunity cost of the discount they are about to give. Yield improvement of even 2 to 4 percent flows almost entirely to operating profit, which is why this is quietly one of the highest-leverage projects in the industry.