Underneath almost every freight commercial operation sits the same artifact: the rate card. A spreadsheet, usually. Sometimes a PDF from a carrier that someone retyped into a spreadsheet. Occasionally a rate module inside a TMS that everyone quietly stopped updating.
The rate card is not a bad tool. It was built for a world where freight moved point to point, on one mode, under one contract, and the commercial question was "what does this lane cost?" Multimodal freight asks a different question, "what does this decision cost?", and rate cards cannot answer it.
The route is not settled
A rate card prices Jebel Ali to Rotterdam, ocean, 40ft. It says nothing about a shipment that is really a factory in Sharjah, a truck leg, a choice between two ports, three carrier options at different transit times, and a rail connection that only works if the vessel arrives before Thursday.
Once a shipment has more than one viable path, price stops being a lookup and becomes an optimization. The cheapest leg on each segment rarely produces the best end to end move, because the legs interact: a $150 saving on the ocean leg that triggers weekend detention at destination is not a saving.
The rate card prices legs. Your P&L experiences journeys.
A quote is a decision, not a lookup.
Send the same request through two quoting systems.
The rate card answers instantly and learns nothing. The live offer prices the moment.
The price is not the price
The base rate is the beginning of the conversation, and every quote is a bet on future costs. Count what the card does not hold on a single Gulf to Europe move: the fuel surcharge that reset this month, the peak surcharge announced last week, in June 2026, carriers announced peak surcharges of up to $2,000 per 40ft container with roughly two weeks' notice, currency adjustments, terminal handling at both ends, and the detention clock that starts when the container clears the gate.
The consequences are measurable. Audit firms consistently find errors on 5 to 10 percent of freight invoices, and systematic audits recover 1 to 5 percent of total freight spend: money quoted at one number and invoiced at another. The people pricing these jobs are not careless. The artifact simply has nowhere to put half the cost structure.
The market has moved
A rate card is a photograph. Freight is a film.
Ocean spot rates roughly doubled during the Red Sea diversions, Drewry's composite passed $5,000 per 40ft in mid 2024, nearly twice its ten year average, then gave the entire gain back by late 2025. Quote from a photograph and you either price above the market and lose the business, or below your real cost and win business you will regret. Both failures are invisible at the moment of quoting.
Most organizations know all this, and compensate with people: the pricing manager who knows the lanes, the WhatsApp thread with three carrier contacts. That knowledge is real. It is not infrastructure. It does not survive scale, and it does not survive turnover.
Manage the decision, not the rate
"Rate management" is the wrong name for the category. Rates are inputs. The thing that needs managing is the decision, and the industry has been buying tools for the input while the decision goes unsupported. Four properties define the alternative:
Price journeys, not legs. Price every feasible routing and show the trade offs, cost against transit time against reliability. The unit of pricing is the option, not the lane.
Model costs, do not list them. Surcharges, commitments, and validity windows live as rules that fire when conditions are met, so the quoted number is the invoiced number.
Build the market in. Contract rates sit alongside live signals: not just "what is our rate" but "where is it, relative to the market, right now."
Make memory the default. Every quote, win, and loss feeds back in. Your best pricing manager's judgment accretes into the system instead of walking out the door.
Integration comes before modeling
In practice this is an integration problem before it is a modeling one. The rates and costs live scattered across carrier platforms, rate marketplaces, and a company's own internal systems, and a pricing engine is only as good as the feeds underneath it. This is the distinction we have drawn between quoting and pricing, applied to the whole multimodal decision.
Build in that order
Our own pricing work is built in that order: integrations that bring external platforms and internal systems into one model, pricing models layered on top, and business parameters, margin rules, commitments, validity windows, governing what goes out the door.
Where AI earns its place
AI earns its place doing what no pricing desk can at quote speed: searching the full option space and returning ranked suggestions in seconds, with the parameters deciding what is allowed to leave the building. The rate card becomes an output of the system, not its source of truth.
Across twenty-two production integrations in seven countries, the pattern has been unmistakable: the companies winning in multimodal freight are not the ones with the best rates. They are the ones who can see, at the moment of decision, what their rates actually mean.
Measure your gap
Pull twenty recent multimodal quotes and compare quoted to invoiced cost, leg by leg. Low single digits: healthier than most, since the audit benchmark says errors alone run 5 to 10 percent of invoices. Anything more, and you now know what your rate card has been costing you.
The question was never "what does this lane cost?" It was always "what is the best decision available right now?" It is time the tools answered it.