Convoy

» see convoy.com as it lived, on the wayback machine

~ the obituary ~

convoy was the kind of us startup you would only encounter if you were in us logistics, us venture press, or us long-haul trucking. i was not in any of those. the first time i heard the name was reading a jassy-cost-cutting-adjacent techcrunch piece in late 2023 that used convoy as the example of what happens when the freight cycle turns after the money is already in.

the company was founded in april 2015 in seattle by dan lewis, an amazon operations manager who had watched the long-haul trucking market up close and thought the broker layer between shippers and independent carriers was ripe for a rewrite. the pitch was simple. instead of thousands of small brokers matching truckers to loads by phone and fax, convoy would run the marketplace as software. shippers post loads and truckers bid through the app. the load routes to whoever accepts first at the price the system suggests.

that pitch got funded aggressively. jeff bezos put in early. bill gates joined the cap table. t rowe price and generation investment came later. the last round in april 2022 was $260 million at a $3.8 billion post-money valuation. eighteen months later dan lewis sent staff an internal memo. the company had spent four months exploring a sale. the freight recession had made it impossible. as of that afternoon, convoy was done.

~ the case file ~

~ the climb ~

the digital freight brokerage thesis was that the incumbent brokers, ch robinson at the top of the pile, were sitting on margin that software could eat. convoy was not alone here. uber freight was in the same segment. transfix, loadsmart, and drayage-focused shops all raised big rounds through 2019 and 2020. the pandemic-era freight boom made everyone look right. spot rates spiked, capacity was tight, and the digital marketplaces were absorbing volume the phone brokers could not price fast enough.

convoy grew fast in that window. by 2022 it was moving loads for anheuser-busch, unilever, procter and gamble, land o'lakes. the pitch to shippers was cleaner pricing and load visibility. the pitch to truckers was you get paid faster and see all available loads in one app. by early 2022 the company was preparing an ipo. the april 2022 round was talked about as pre-ipo capital.

~ how it actually ended ~

the freight cycle turned in the second half of 2022. spot rates collapsed as pandemic inventory backlogs cleared and consumer spending shifted. by mid-2023 the us truckload spot market was down 40 percent year over year on average. convoy's unit economics assumed a certain floor on spot pricing. that floor was gone. the burn rate did not shrink fast enough to match.

dan lewis's october 18 memo laid it out plainly. the board had authorized a sale process in june. four months of buyer conversations produced no viable offer. the freight recession combined with a capital-markets contraction that made new equity rounds impossible. existing debt covenants left the company without a runway. operations halted the next day. 533 of roughly 600 remaining employees were let go.

several carriers reported that convoy owed them for completed loads at shutdown. the amount was contested. flexport acquired the technology and intellectual property in a small deal announced in november 2023. the convoy brand was later reused by flexport under a lighter-weight version of the platform. the original company was gone.

~ the long tail ~

convoy is the anchor of the post-zirp unicorn-collapse cluster that runs through 2023 to 2025. the freight recession broke the marketplace layer at the same time as bowery farming's collapse broke the vertical-farming layer and forward health broke the clinic-chain layer. seven or eight companies at similar scale went down in the same window.

the specific lesson from convoy that logistics press keeps writing about is that a real-world marketplace with high unit costs and cyclical pricing is a bad match for a growth-stage cap table. software marketplaces at ebay scale work because the unit costs are near zero. freight brokerage does not have that shape. by 2026 the digital brokerages that survived did so by pairing the software with owned assets or by staying much smaller than convoy tried to.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »