Race to Build Dark Store Footprints
Former head of strategy at a global on-demand giant on the economics of grocery delivery
This phase rewarded balance sheet speed more than customer love. In ultrafast grocery, each new city meant leasing a dark store, stocking inventory, hiring pickers and couriers, and getting enough neighborhood coverage to promise 10 to 20 minute delivery. Because many stores were still doing only 30 to 50 orders a day, the scarce asset was not demand efficiency yet, it was the ability to plant supply fast enough to matter before rivals filled the map.
-
A dark store was cheap enough to copy city by city, roughly £100,000 to open, which turned expansion into a capital deployment contest. The companies with the most funding could open more nodes, cover more neighborhoods, and learn operations faster, even before proving steady customer pull.
-
This was different from the older marketplace model used by Uber, Deliveroo, and Glovo, which piggybacked on existing grocers and courier fleets. Dark stores required owning inventory and fulfillment, which raised fixed costs but gave tighter control over speed, assortment, and delivery consistency.
-
The logic was that footprint comes first, then demand. Once a network exists, operators can shift into customer acquisition, bigger baskets, and better economics. That is why early leaders looked more like infrastructure builders than marketers, and why consolidation was expected once enough of the map had been claimed.
The market then moves from land grab to density game. The winners are the ones that can turn a wide but underused network into busy neighborhood stores with repeat orders, larger baskets, and better delivery batching. Fast grocery stops being a funding contest when each dark store has to earn its keep.