Courier Networks as Delivery Infrastructure

This reveals that delivery marketplaces become much better businesses when they sell logistics as infrastructure instead of buying demand on both sides of the market. In the full marketplace model, DoorDash and Uber have to attract diners, sign restaurants, and keep enough drivers online. In Drive and Direct, a partner like ChowNow already brings the restaurant and the order, so the courier network can be sold more like a metered utility with flatter, more predictable unit economics.

• For a restaurant software company, this setup removes the hardest part of delivery, finding and dispatching a driver at the right moment. The restaurant keeps the customer relationship on its own site or app, while DoorDash or Uber just handles pickup, live tracking, and dropoff for a per delivery fee.
  1 sacra 2 sacra 4 doordash 5 uber
• The economic gap is large. Full marketplace delivery often took roughly 20% to 30% or more of an order, while the restaurant stack that mixed subscription software with third party delivery infrastructure came in near an 11% blended take rate, with DoorDash Drive described in prior research as roughly $7 to $11 per order.
  2 sacra 3 sacra 5 uber 6 sacra
• This also creates a strange coopetition dynamic. DoorDash and Uber compete with ChowNow, Owner, and Lunchbox for restaurant relationships, but they also power those same companies behind the scenes. That makes the courier network more valuable because it can monetize orders that never happen inside the marketplace app.
  1 sacra 3 sacra 4 doordash 6 sacra

The next phase is that courier networks keep moving down the stack and become default delivery rails for restaurant software, retail checkout flows, and local commerce more broadly. As that happens, the winners in software will own the merchant and customer relationship, while the winners in logistics will own density, routing, and driver supply across as many order sources as possible.