Ryan Petersen: Building Flexport into a $3B Logistics Giant

KEYPOINTS

• 01

  In logistics, quality costs less.

  Petersen calls it counterintuitive but central: almost all costs come from mistakes. File a customs entry with the wrong classification code and you can spend weeks with the agencies, or face licence and regulatory trouble. One error can undo a month of efficiency gains, so getting it right first time is the cheapest path. He traces the idea to W. Edwards Deming.

• 02

  Over-optimising for efficiency broke Flexport's service.

  Flexport split shipments into tiny Amazon-style tasks, so an operator did one or two steps repeatedly and 50 people might touch one shipment. That suits fulfilment centres, not freight forwarding, where truckers no-show, APIs go silent and cargo gets lost. Net promoter score, long in the 70s, fell to about 17. The fix was a reorg back to Toyota-style ownership: one operator follows a shipment end to end.

• 03

  Stepping down because someone else would be a better CEO was a mistake.

  Seeing himself as a 'freewheeling entrepreneur' rather than an operator, Petersen hired Amazon's Dave Clark, first as co-CEO, then as sole CEO for six months. With $1.6 billion in the bank, they leaned in while others cut, adding about 900 mostly US engineers in 12 months as freight and capital markets collapsed. He takes the blame for approving it. Paul Graham's reaction: 'That's like saying this other guy would be a better husband for your wife.'

• 04

  Micromanagement is just attention to detail with bad PR.

  Advice from Tobi Lütke and Brian Chesky changed how he ran the company on his return. Don't treat senior hires as black boxes; audit, inspect, do skip-levels. He talks to 40–50 people a day across Slack, text and phone, then brings issues back to their managers. Chesky's line gave him confidence: he knows he'll be there in ten years, and he can't be sure anyone else will.

• 05

  Promote from within, and give outside hires a 90-day no-decisions rule.

  He has made just one outside hire in two years and told him he couldn't make decisions for 90 days. On day 91, the hire said he'd have decided differently with only 80. Executives joining a company should bring extreme humility; teams have obsessed over these problems for a decade.

• 06

  Choose your bottleneck or it chooses you.

  In a healthy B2B company, the constraint should be customer demand. Flexport's trade-advisory team became an unplanned bottleneck as tariffs required far more expertise. Amazon puts the bottleneck at the most capital-intensive machine and overbuilds it — though Petersen thinks it pushed the jam onto truck unloading.

• 07

  Judge policies by outcomes, not intentions — and the tariffs are backfiring.

  He has met no customer onshoring because of the tariffs, but at least half a dozen moving production out. A US kids'-bike maker paid duties on components while finished bikes came in duty-free. Exporters assembling in the US pay duties on inputs, and a friend's German-made chip-fab machinery would have cost 30% more. When duties on China hit 145% on 9 April, volumes fell 60%; when cut to 30%, they surged 80% above pre-tariff levels.

• 08

  Retention sets your ceiling.

  At Import Genius, steady signups met high churn — users ran a search and left. Your equilibrium size is new signups divided by churn rate. Spending to bend that curve mostly added cost. The loyal customers were freight forwarders building lead lists and hedge funds.