Rent and acquisition as unit economics
David Lin, CEO of Duffl, on the economics of hyperlocal ultrafast delivery
The real bottleneck in ultrafast delivery is not just picking and dropping off an order, it is paying for the small physical network and local demand engine that make 10 minute delivery possible. Duffl treats a store lease near campus and the cost to get students to open the app as part of the cost of each order, because those costs are what create order density, and density is what lets a rider stack trips, lower labor cost, and make a small basket profitable over repeated weekly purchases.
-
In the broader ultrafast market, contribution margin is often shown after cost of goods, labor, and delivery, with rent and acquisition left outside the frame. The more complete store by store view includes both, because every new zone needs nearby space and a steady stream of local customers before fixed operations can be absorbed.
-
Duffl’s campus model is built to compress both of those costs. Dense campuses let one racer handle 10 to 12 orders per hour, versus 3 to 6 for many peers, and 80% of surveyed customers said they discovered Duffl through social connections or referrals, which lowers paid acquisition needs.
-
That is a sharp contrast with larger dark store players like Gopuff, whose model scaled through 200 plus fulfillment warehouses and heavy staffing, creating long payback periods and large losses. Bigger footprints can buy assortment and reach, but they make rent and utilization much harder to ignore in per order economics.
Going forward, the winners in ultrafast delivery will be the operators that make each neighborhood or campus work as a tiny self contained profit center. That means faster store payback, tighter real estate, more organic demand, and higher repeat frequency, not just lower picking and delivery cost on a spreadsheet.