Freight Brokerage and Digital Freight Platforms

US trucking is the most fragmented large freight market in the world — hundreds of thousands of carriers, most running fewer than ten trucks — and the freight broker is the market-maker that connects it to shippers. A broker owns no trucks: it wins a load from a shipper, finds a carrier to haul it, pays the carrier less than it charges the shipper, and keeps the spread (its gross margin, or net revenue). Brokers therefore report large gross revenue that is mostly the carriers' money: C.H. Robinson's North American Surface Transportation segment reported Q3 2026 revenue of ~$3.83bn but adjusted gross profit of ~$518m (company 8-K, 28 Oct 2026). Brokerage margins move against the spot market: when spot truckload rates fall below the contract rates brokers have agreed with shippers (2019, 2022–25), margins widen; when spot spikes above contract (2018, 2021, and the supply-driven tightening of 2026), margins compress — unless the broker reprices, as C.H. Robinson did in 2026, lifting truckload adjusted gross profit per shipment ~17% y/y while average headcount fell ~10% (company). The 2017–22 digital-freight boom poured billions of venture dollars into app-based brokers promising to replace the phone and the broker rep with algorithms; most failed or retreated when the freight recession arrived — Convoy, once valued at ~$3.8bn (2022, reported), shut down on 19 Oct 2023; Transfix sold its brokerage operations (2024, reported; verify); Uber Freight survived by buying the managed-transportation firm Transplace (2021, ~$2.25bn). Incumbents absorbed the technology instead, and from 2023 the frontier moved to AI agents that quote, book, track and collect. The industry's other story is fraud: identity theft of carrier authority, double brokering and fictitious pickups rose sharply in 2022–26, while FMCSA told Congress it lacked data even to measure double brokering and asked for authority to fine unauthorised brokers (reported). This section explains the broker's model, gross versus net revenue, margins through the cycle, the main brokers (C.H. Robinson, TQL, RXO, Landstar's agent model), digital freight's rise and fall, load boards, fraud and regulation, liability and AI. It has no worked example: the arithmetic of one broker's per-load economics and working capital is worked in Appendix A.4 of the main book, and an owner-operator's cost per mile through the cycle — the other side of the broker's spread — in 04.