Dark Stores Capex Payback Crisis
Instacart vs Amazon vs Uber
The dark store model broke when growth had to pay for the warehouses it had already built. Unlike Instacart, Uber, or DoorDash, which could route orders through existing stores and courier networks, ultrafast players had to lease sites, buy inventory, hire pickers, and keep each neighborhood hub busy enough to cover fixed costs. When funding tightened in 2022, that gap between upfront buildout and later payback turned into layoffs, closures, and consolidation.
-
A dark warehouse is not just software and drivers. It is a mini store, usually around 3,000 square feet, stocked in advance with only a few thousand SKUs, staffed for picking, and placed close enough to customers to hit 10 to 15 minute delivery times. That speed advantage came with rent, inventory, and labor before demand was proven.
-
The payback math depended on very high order density. In a mature dark store doing about 500 orders per day, contribution margin could look positive, but that excluded overhead like customer acquisition and some occupancy costs. If a location ramped slower than planned, every extra warehouse extended the cash burn instead of improving economics.
-
By 2022 the market stopped funding that burn. Gopuff cut about 10% of its workforce and closed warehouses to preserve cash, while across quick commerce the year brought layoffs, pullbacks, and deals like Getir buying Gorillas. The companies with existing store infrastructure were left in the stronger position.
Going forward, instant delivery keeps moving toward asset lighter models and incumbent infrastructure. The winners are more likely to use existing supermarkets, convenience stores, and courier fleets, then layer in higher margin ads, software, and fintech. Building a new warehouse network from scratch now looks less like an advantage and more like a financing risk.