Convoy ran out of road. Its software kept going.

An empty truck still burns diesel. It still occupies a driver’s day. It simply carries nothing anyone will pay to move. Convoy built its proposition around that awkward fact: trucking had plenty of work and plenty of trucks, yet connecting the two required an astonishing amount of calling, waiting and guessing. Put that coordination into software, the founders reasoned, and everybody could keep more of the value.

• Then: a digital brokerage moving freight for large shippers.
• The break: the core business closed in October 2023.
• Now: a DAT-owned marketplace helping brokers and carriers execute loads.

The idea produced a company valued at $3.8 billion in April 2022. Eighteen months later, its core brokerage closed. Yet the technology kept finding buyers. Flexport acquired the assets, relaunched the platform, then sold it to DAT in July 2025. Convoy’s strange achievement was to make something useful enough to outlive the organization that built it.

First, go where the trucks are

Dan Lewis and Grant Goodale, both Amazon alumni, founded Convoy in Seattle in 2015. Before the launch, Lewis spent weeks talking to people in warehouses and truck stops. The questions were wonderfully unglamorous: what makes a shipper’s day difficult? What makes a carrier’s day difficult? A marketplace needs both answers. Solving one side’s inconvenience by handing it to the other is a fragile business.

The resulting product let shippers obtain prices and arrange transportation while trucking companies found work through an app. Convoy earned money from the transaction, keeping a portion of the freight payment. Its automation aimed to make a brokerage less labor-intensive. Customers included Unilever, Anheuser-Busch and Procter & Gamble, businesses with goods that needed actual trailers, appointments and delivery confirmations.

The technical expertise lay in pricing, matching and execution: which carrier could take a shipment, at what rate, with what likelihood of completing it? Traditional brokers and rivals such as Uber Freight were pursuing the same customers. Convoy’s wager was that software could handle more of the routine coordination. In February 2019, it reported automated matching for 95% of loads nationwide. Matching, however, was only one part of running the business.

The next load mattered, too

Automated Reloads, introduced in 2019, tried to solve a more interesting problem than finding one job. It grouped compatible shipments so carriers could book a sequence. The next pickup had to make sense after the previous delivery, considering location, waiting time and driving hours. A bargain load followed by a long empty drive could be a poor bargain for the carrier.

Convoy also introduced Go, a drop-and-hook program built around preloaded trailers. That extended the proposition beyond an app: carriers could collect ready trailers rather than wait through loading. The original company’s culture encouraged contact with these operational details. One leadership retreat took employees to a truck stop for ride-alongs. New hires went through product demonstrations for both sides of the marketplace.

“We want to build cars, not faster horses.”Grant Goodale, 2017

There was a catch in that ambition. A trailer network brought commitments that a matching algorithm alone did not. Looking back in 2024, Lewis described the late-2019 expansion into trailer pools and contracts as an expensive bet when the pandemic changed freight conditions. The system could become more capable while the company became harder to finance.

The capital stopped before the code

Convoy’s April 2022 announcement separated three different kinds of money: $160 million in equity, $100 million in venture debt and a $150 million credit facility. The distinctions matter. Borrowing capacity is not revenue, and a valuation does not pay an invoice. Convoy still needed a sustainable relationship between what shipments earned and what the organization cost.

In his shutdown memo, Lewis pointed to collapsing freight conditions and tighter capital markets. Cost cuts had not made the business sufficiently self-supporting; financing and strategic alternatives had not come together. The immediate failure was the company’s ability to keep operating. The public record does not establish that its matching software had suddenly stopped working.

Lewis later acknowledged that the commitment to the whole system had made it difficult to separate its pieces. That is the uncomfortable management question here: when customers value a capability, must the company own every surrounding activity to deliver it? Convoy’s successors answered by drawing a smaller boundary around the product.

Sell the machinery to the brokers

Flexport explicitly acquired the technology rather than Convoy the corporation or its liabilities. Under the relaunched model, other brokers could use the platform. DAT subsequently acquired it. The current website identifies DAT Freight Tech, LLC as operator and says the platform has no brokerage authority. Brokers keep their customer relationships and remain responsible for fulfillment.

The workflow now connects a broker’s transportation management system to carriers. A broker provides a load and spending constraints; carriers bid or, when offered, book at a posted rate. Tracking, paperwork and payment follow. For the carrier, the appeal is fewer separate administrative steps. For the broker, it is more capacity and less repetitive coordination.

Carrier access has no signup or monthly fee. Standard 30-day payment is free. Optional two-day QuickPay costs 1% with the qualifying app-use discount, versus 2% without it, subject to applicable minimums. Same-day options vary by broker. Broker platform charges follow agreed terms; DAT’s schedule describes shipment and percentage-based fees. Freight rates themselves vary by load.

A smaller promise, with useful limits

This works best when the freight fits the network. The current broker offering supports single-driver full truckloads with specified equipment and live loading. It does not make every specialist shipment eligible. Nor can software guarantee a convenient return load where demand, timing and truck location fail to overlap. Exceptions still require people, and brokers retain responsibilities after booking.

The practice worth copying is specific: interview the people on each side, automate work that repeats, and test whether the resulting tool can serve existing operators. Convoy’s original ambition needed the company to coordinate much of trucking itself. Its surviving product gives more people the machinery to coordinate their own corner of it.