Everyone remembers January’s silver spike. Then what?

The storefront is still there. The address is the same. But the lights are off, the aisles are rearranged for picking speed rather than browsing, and no customer will ever walk through the door again. This is not a story about retail failure. It is a story about deliberate conversion, and it is accelerating faster than most real estate investors have priced in.

Credible forecasts do not support a single clean set of numbers for the global dark store market in 2025 to 2026, but they do agree on the direction: rapid growth. For example, Grand View Research estimates the global dark store market at about $15.27 billion in 2023 and projects roughly $129.25 billion by 2030. Other market trackers cluster in the same broad range for 2030, but the point is consistent: the category is scaling from “pilot” to “infrastructure.” Those figures are not driven by niche operators. Walmart, Amazon, Kroger, and Target are commonly cited as major competitive forces reshaping this segment right now.

Amazon’s move is the sharpest signal. In May 2026, Amazon rolled out its 30-minute delivery service, Amazon Now, making it widely available in Atlanta, Dallas-Fort Worth, Philadelphia, and Seattle, with availability also expanding into parts of dozens more cities including Austin, Denver, Minneapolis, and Phoenix. Prime members pay $3.99 per order, with an additional small-order fee on carts under $15. The infrastructure behind that promise is a network of smaller, local fulfillment nodes placed close enough to reduce the last-mile cost that destroys margin at longer distances.

The labor math is unambiguous. A picker navigating a live supermarket aisle assembles roughly 60 to 80 units per hour. A dark store can push that to about 120 to 150 units per hour, and more automated micro-fulfillment systems can run faster still. That gap, multiplied across thousands of weekly orders, is the difference between a grocery delivery operation that bleeds margin and one that survives. Grocers that treated delivery as a bolt-on to existing stores encountered a structural ceiling: in-store picking clogs aisles, competes with walk-in customers, and caps throughput at a level that cannot support same-hour economics.

The commercial real estate consequence is measurable, but it is not uniform by market. Research firms tracking industrial fundamentals have repeatedly flagged small-bay infill space as an outlier: infill warehouses roughly in the 10,000 to 50,000 square foot range have been running at mid-single-digit vacancy rates, materially tighter than larger bulk industrial. Vacant big-box footprints are being converted into fulfillment hubs as landlords chase the tenant category that is still expanding its square footage requirements. A single automated MFC is commonly cited in the low single-digit millions up to the high single-digit millions in capital cost, which means operators with real balance sheets are the ones signing leases, not marginal tenants.

The model does have a structural ceiling of its own. Quick-commerce dark stores often carry roughly 1,500 to 4,000 SKUs, a fraction of a full supermarket’s assortment. When orders spike unexpectedly, a constrained node cannot absorb the volume. Scaling requires adding locations rather than expanding any single footprint, which compounds real estate costs across dense urban grids where the best sites are already spoken for.

What this shift means for retail real estate is not a simple vacancy crisis. It is a reclassification. The question property assessors, zoning boards, and REIT managers have not yet answered cleanly: is a micro-fulfillment center inside a former supermarket industrial space or retail space? That classification determines tax treatment, permitted use, and lease comparables. The market has moved faster than the regulatory framework has followed, and that gap is where near-term risk concentrates for investors holding legacy retail assets in dense urban corridors.