For years, transportation technology has treated rating as a solved problem.
Enter a rate. Apply a markup. Calculate the shipment cost. Move on.
Nearly every TMS can handle the basics.
But transportation has become far more complex and if your rate engine can’t handle the complexities, issues begin to arise.
Customers expect more specialized services. Carriers operate under increasingly varied pricing structures. Brokers and 3PLs are managing more modes, more accessorials, more contractual requirements, and more exceptions.
The result is an uncomfortable reality:
The industry doesn’t have a shortage of rate engines. It has a shortage of rating infrastructure that can keep up with the complexity of modern freight.
When “rating” becomes engineering
Weight sounds straightforward until it isn’t.
Dimensional rating may require comparing actual weight against dimensional weight and applying whichever is greater. Then the same customer may want the calculation performed using a different unit of measure.
Multiply that across pallets, boxes, bags, drums, handling-unit breaks, and customer-specific rules, and “apply the rate” becomes something very different.
The same is true of geography.
A simple ZIP-code-based rate can work well for a straightforward shipment. But what happens when a truckload has four stops spanning three regions?
Which geography determines the rate?
How are the intermediate stops treated?
What happens when distance, weight, equipment, dates, and handling units all influence the answer?
There isn’t always one universally correct calculation.
There is only the calculation defined by the business.
And when a system can’t represent that logic, the complexity doesn’t disappear. It moves into spreadsheets, workarounds, and employee expertise.
The hidden infrastructure behind transportation pricing
This is where an important distinction emerges.
A rate is data. Rating is logic.
Modern transportation businesses aren’t simply storing thousands of rates. They’re applying rules to determine which rate should be used, when it should be used, and how it should interact with everything else involved in the shipment.
Consider just a few common scenarios:
- Storage priced according to both volume and time
- Accessorials triggered by specific shipment conditions
- Taxes calculated against freight, fuel, and additional charges
- Rates that change based on weight or handling-unit breaks
- Multi-stop shipments with geography-specific rules
- Customer-specific pricing agreements
- Carrier rates with different calculation methods
- Charges that depend on dates, service levels, or equipment
None of these are unusual edge cases.
They’re part of operating transportation businesses today.
Yet many systems still approach rating as fundamentally a rate-table problem.
That’s where the gap starts to show.
The spreadsheet is telling us something
Consider a final-mile operation with hundreds or even thousands of carriers.
Each carrier may have its own pricing model. Some charge per mile. Others use weight breaks. Others price by stop, hour, piece, or some combination.
Eventually, someone has to normalize all of those structures to determine what a shipment actually costs.
Too often, that someone is a person with a spreadsheet.
The spreadsheet isn’t the problem.
The spreadsheet is evidence of a system architecture problem.
It represents business logic that the core technology can’t easily express.
And that creates three consequences.
First, pricing becomes dependent on individual expertise.
Second, scaling becomes increasingly difficult.
Third, the organization has trouble seeing its true economics in real time.
A business can have sophisticated pricing talent and still struggle to turn that expertise into a repeatable, scalable process.
Normalization changes the conversation
Once different pricing structures can be represented and normalized consistently, something more interesting happens.
The question stops being:
“Who is cheapest?”
It becomes:
“Who is the best option for this shipment?”
A carrier that costs slightly more but delivers significantly better on-time performance may be the better economic choice.
A carrier with stronger claims performance may reduce downstream costs.
A carrier with the right insurance coverage may be the only viable option for a particular customer.
A carrier with better service reliability may protect the customer relationship.
Price is still important.
But price is only one variable in the decision.
That means the next generation of transportation technology needs to connect rating with the broader decision-making process: carrier qualification, service performance, risk, margin, customer commitments, and operational constraints.
Rating should inform the decision, not simply produce a number.
This matters even more in the age of AI
The transportation industry is increasingly asking what AI can do.
But AI doesn’t eliminate the need for accurate business logic. It makes that logic more important.
An AI system can only make a reliable recommendation if it has access to the right data and understands the rules governing the decision.
If the carrier agreement lives in a PDF, the customer exception lives in an email, the pricing logic lives in a spreadsheet, and the operational workaround lives in someone’s head, the AI doesn’t have a complete picture.
It may be intelligent.
It just doesn’t have the information required to make the right decision.
This is why the future of transportation automation isn’t simply about putting AI on top of existing systems.
It is about making the underlying business logic accessible, structured, and executable.
Rating is one of the clearest examples of this problem.
From rate engine to decision engine
The transportation technology market has spent years competing over features.
But the more important question may be architectural:
Can the system represent the complexity of the business?
Can it handle multiple transportation modes without creating separate islands of logic?
Can it represent customer-specific agreements?
Can it normalize fundamentally different carrier pricing structures?
Can it apply qualification rules in a defined hierarchy?
Can it incorporate service, risk, and margin into a decision?
Can that logic be exposed through APIs so it can power existing systems, new applications, and AI-driven workflows?
If the answer is no, adding another feature isn’t going to solve the underlying problem.
The industry doesn’t need another checkbox labeled “rate engine.”
It needs technology that can turn complex transportation economics into structured, repeatable, executable logic.
Because having a rate engine was never the achievement.
The real achievement is making transportation complexity manageable at scale.
And that is where the next generation of freight technology will differentiate itself.