US business logistics costs reached $2.4 trillion in the most recent accounting, or 7.8 percent of GDP, according to the Council of Supply Chain Management Professionals in its 37th annual State of Logistics Report, produced by Kearney and presented by Penske Logistics. A meaningful share of that figure passes through firms that own no trucks and no warehouses. Armstrong and Associates' market size estimates put non-asset domestic transportation management, the segment where freight brokerage sits, at $128.3 billion of gross revenue in 2025, within a US third-party logistics market of $323.4 billion. The work these firms do is the same kind of work a manufacturer does at month-end or a distributor does in its order desk: a high volume of small, repetitive transactions, each one requiring documents to be read, statuses to be communicated, and invoices to be checked. Freight is one instance of a pattern that recurs across operating businesses, and it is a useful instance because the margins are thin enough that the cost of each manual touch is easy to see.
What a load actually requires
A single shipment generates a sequence of administrative steps that have little to do with the broker's actual skill, which is knowing carriers and pricing capacity. The load is quoted and booked. A carrier is sourced and tendered. The shipment is tracked from pickup to delivery, with the customer expecting updates along the way. Documents arrive: a rate confirmation, a bill of lading, a proof of delivery, sometimes accessorial paperwork for detention or a lumper fee. The carrier sends an invoice, which the broker audits against the agreed rate and the documents before paying, and the broker in turn invoices the customer. Most of these steps are coordination and data handling rather than judgment. A broker handling hundreds of loads a week is doing the same document extraction and the same status communication hundreds of times, by hand, with a person reading each PDF and typing into a transportation management system.
Reading the documents
The first place a redesign starts is the inbound document, because that is where the manual typing concentrates. A rate confirmation, a bill of lading, and a proof of delivery are structured in the sense that they contain the same fields every time, and unstructured in the sense that every carrier formats them differently and many arrive as scans or emailed photos. An agent extracts the load number, the rate, the weight, the pickup and delivery details, and the accessorial charges, then writes them back into the system of record. The difficulty is not the typical document. It is the long tail of formats and damaged scans, which is why the redesign separates the documents the agent can read with confidence from the ones a person should look at. We have written separately about the long tail of document variability and why it, rather than the common case, governs how a data-entry process should be rebuilt.
Status updates without the check call
The second concentration of manual effort is the status update. Customers want to know where their freight is, and in a traditional brokerage that means a coordinator making check calls to drivers and relaying answers. A redesigned process pulls location and milestone data from tracking sources and generates the customer-facing update directly, on a schedule the customer cares about, so the proactive notification goes out without a person assembling it. The carrier relationship and the customer relationship stay with the broker's people, where they belong, because those relationships are the asset. What changes is that the routine "where is my load" exchange no longer consumes a coordinator's afternoon.
Auditing the freight bill
The third area is the one where margin makes the case most directly. Carrier invoices do not always match the agreed rate. They include charges that were never authorized, duplicate a line, or apply an accessorial that the documents do not support. In an industry where the broker's spread is narrow, an overpayment that survives because no one checked the bill is a direct subtraction from the only money the broker keeps. An agent compares each carrier invoice against the rate confirmation and the supporting documents, approves the ones that tie, and flags the discrepancies for a person to decide. This is the same logic as three-way match, rebuilt in accounts payable, where the machine does the mechanical comparison across documents and a person adjudicates the exceptions.
Why the thin margin changes the math
In a business with wide margins, a few dollars of leakage per transaction is absorbed by the spread and no one builds a process to catch it. Freight brokerage does not have that cushion. The US truckload sector spent years in a freight recession, and CSCMP's 2024 State of Logistics Report counted more than 1,000 US freight brokers that closed in the year after its 2023 edition. When rates are compressed, the difference between a broker that audits every bill and writes every status update by hand and one that has redesigned those steps shows up in whether the firm survives the cycle. Transaction-level efficiency is consequential here for the same reason it is consequential anywhere the unit economics are tight: the cost of the manual touch is a larger fraction of what each transaction earns. This is the same calculation that drives process redesign in a wholesale distributor, where order-desk volume meets a similar margin structure.
The relationship stays human
A reasonable counter is that brokerage is fundamentally a relationship business, that customers and carriers stay with a broker because of the people who answer the phone and solve problems, and that automating the back office risks the very thing that wins the business. There is real truth in that, and a redesign that pushed agents into the carrier negotiation or the customer escalation would be making a mistake. But the relationship is not built on a coordinator retyping a bill of lading or reading a tracking page back to a caller. It is built on availability, pricing, and problem-solving, and those are exactly the activities the people keep when the document handling, the routine status update, and the invoice audit move off their desks. The redesign returns time to the relationship work instead of taking it away, which is the general shape of what the operator's role looks like after automation: less mechanical handling, more of the judgment and the contact a machine cannot do. The agent runs against the broker's existing transportation management system, writing back where it can and routing what it cannot, so the firm keeps its systems, its carriers, and its customers, and reclaims the hours that the thin margin makes count.