Swiggy's quick-commerce arm Instamart added 69 net dark stores in the September quarter to reach 1,240, while Eternal's Blinkit added 308 over the same 92 days to reach 2,751. As of 5 November 2026, the difference between the two networks is no longer ambition but unit economics: Blinkit reported adjusted core profit of Rs 1.84 billion, while Instamart's adjusted core loss widened to Rs 8.95 billion from Rs 8.49 billion a year earlier. For brands that manage distribution and on-shelf availability, store counts are becoming a misleading headline.
Instamart stops widening and starts deepening
The strategic shift is explicit in Swiggy's own language: the company said it would focus on deepening Instamart's presence in existing markets rather than expanding into new ones, a measured approach after several quarters of aggressive build-out. Jefferies, commenting on the share price reaction, argued that Swiggy's valuation does not fully reflect the food-delivery business and its cash holdings. Both statements point at the same thing. The market has stopped paying for footprint and started asking what each store earns.
That is a change in the operating question. For four years the sector competed on how fast a network could be built and how many postal codes it could claim. Now the same operators are being asked how many orders a single dark store needs per day to cover rent, pickers, riders and shrinkage, and how much of that volume comes from customers who would have bought anyway through a supermarket. The answers decide which of the two expansion paths survives the next funding cycle.
What 1,240 stores buys and what it does not
Store count is a proxy for coverage, not for availability. Two networks of similar size can behave very differently depending on how stores are distributed inside a city: a cluster of small stores in dense neighbourhoods serves short delivery radii and fast picking, while a smaller number of larger sites lowers rent per square foot but lengthens the last mile. Instamart's 1,171 stores across 131 cities in June and Blinkit's 2,443 in June describe two different geometries, not just two different totals.
| Operator | Dark stores | Net adds in quarter | Adjusted core result |
|---|---|---|---|
| Blinkit (Eternal) | 2,751 in September 2026 | +308 | Profit of Rs 1.84 billion |
| Instamart (Swiggy) | 1,240 in September 2026 | +69 | Loss of Rs 8.95 billion |
| Zepto | 1,139 across 66 cities in March 2026 | Not disclosed here | Not disclosed here |
Read the table vertically rather than horizontally. Blinkit added more stores in one quarter than Instamart operates in a tenth of its cities, and it did so while turning its adjusted core result positive. That combination is only possible if the marginal store is profitable early, which depends on order density in the catchment rather than on the size of the city. Coverage without density produces the opposite: more fixed cost, longer delivery times and a thinner assortment per store.
Density beats coverage where the cost sits
Quick commerce carries a cost structure that punishes distance. A store has to hold enough assortment to serve most baskets, which means chilled and frozen capacity, and it has to be close enough that a rider can complete several deliveries an hour. Every additional kilometre in the delivery radius reduces the number of trips a rider completes, and rider cost is charged per order rather than per kilometre. Density therefore buys margin twice: through faster picking and through more trips per shift.
The cost line moves from rent to riders
When a network chooses many small stores over few large ones, rent per square foot rises but delivery distance falls. When it chooses the opposite, rent falls and rider cost rises. The crossover point is different in every city because it depends on traffic, rider supply and how much of the assortment is chilled. This is why a single national benchmark for store economics is misleading, and why operators now talk about mature versus new cities instead of one blended number.
Owned inventory changes what revenue means
Eternal's filing notes that Blinkit moved from a marketplace model to one where it owns most of the stock, so it books the full price of the item as revenue rather than only its commission. The customer pays the same; the revenue line counts the whole basket. Any comparison of quick-commerce revenue across operators has to establish which model sits behind the number before drawing conclusions about scale, and the same caution applies to gross margin percentages reported by different platforms.
The fee stack is where the shopper feels the difference
Platform economics are invisible to shoppers, but the fee stack is not. A comparison of the same small basket across the three apps in November 2026 put an Rs 84 order at Rs 143 on Blinkit, Rs 115.50 on Zepto and Rs 154 on Instamart after all fees. Handling fees were Rs 4, zero and Rs 9.80 respectively, with free delivery thresholds at Rs 199, Rs 99 and Rs 199. Small-basket economics, not store count, decide who wins the top-up trip.
| Item | Blinkit | Zepto | Instamart |
|---|---|---|---|
| Handling fee | Rs 4 | None | Rs 9.80 |
| Free delivery from | Rs 199 | Rs 99 | Rs 199 |
| Total on an Rs 84 basket | Rs 143 | Rs 115.50 | Rs 154 |
What brand teams should change in their distribution checks
The practical consequence is that distribution coverage should be measured at store level, not at city level. A brand listed in a city but out of stock in the four stores that serve the densest catchments has, in effect, no availability where it matters. The same logic applies to new launches: a listing that exists in the app but sits behind a store with no chilled capacity will not convert, and the sales report will attribute the failure to demand rather than to distribution.
It also changes how category growth should be read. Platforms report strong consumer demand and double-digit category growth, but the growth is not evenly distributed across stores, and the stores that carry it are the ones with density. A brand planning assortment by city will over-invest in coverage and under-invest in the handful of catchments that generate most of the volume.
FAQ
Does a larger dark-store network mean better product availability?
Not by itself. Availability depends on how stores are distributed inside a city and whether each store is replenished in time. A network with fewer but denser stores can show higher availability than a larger, more dispersed one, which is why store-level checks matter more than national counts.
Why did Swiggy stop entering new cities?
Swiggy said it would focus on deepening Instamart in existing markets rather than expanding into new ones. The stated reason is a more measured approach after aggressive expansion; the financial context is a widening adjusted core loss at Instamart against a positive adjusted core result at Blinkit.
What should a brand watch when a platform changes its inventory model?
Revenue recognition changes before anything the shopper sees. When a platform moves from marketplace to owned inventory it books the full basket value rather than a commission, so revenue growth can jump without a single extra order. Compare order volume and store count alongside revenue.
