As of November 12, 2026, the most useful number in Indian quick commerce is not the store count. Instamart closed the September quarter with 1,240 active dark stores across 133 cities, 5.21 million square feet of space, and 1,192 orders per store per day. It also posted an adjusted operating loss of Rs 895 crore and a contribution margin of negative 1.0 per cent of gross order value. The gap between those two sets of numbers is the whole story: the network grew, the losses grew, and the per-store figure is the only one that tells you whether either is working.
What the quarter actually added
Three figures describe the quarter. Gross order value rose 31 per cent year on year to Rs 9,230 crore. Net order value, which strips out discounts, rose 34.8 per cent to Rs 6,594 crore, and sequential growth accelerated to 13.4 per cent from 2.5 per cent. Monthly transacting users reached 14.8 million, up 23 per cent. On the cost side, the adjusted EBITDA loss widened from Rs 849 crore a year earlier to Rs 895 crore, while the margin as a share of GOV improved to negative 9.7 per cent from negative 12.1 per cent.
That combination is not a contradiction. A business can improve its per-order economics while losing more money in absolute terms, because it is adding fixed capacity faster than it adds orders. The margin tells you the machine is getting more efficient; the loss tells you the machine is getting bigger. Whether that is good news depends entirely on how fast the new capacity fills.
A store opens when the neighbour saturates
The disclosure does not include a per-store threshold for opening a new site, which is the number that would make the expansion legible. What it does include is the shape of the decision: 69 net additions in a single quarter, the same as the previous three quarters combined, in a network that stayed at 133 cities. Adding that many doors inside an unchanged footprint is a density move, not a coverage move. A trade area is no longer scored by how many households sit within two kilometres, but by whether the nearest store is already saturated and whether a new one can take volume without starving it.
The same logic explains why the average store does 1,192 orders a day while 5.21 million square feet sits behind it. Some stores are mature and comfortably above the average, many are new and well below, and the mean lands in between. Read on its own, 1,192 looks like a modest utilisation figure. Read against the fact that a quarter of the network is brand new, it says the network is carrying a block of capacity that has not yet filled.
Cannibalisation is the number nobody publishes
When a new store opens 700 metres from an existing one, the old store's volume can drop sharply while its rent, staff and equipment costs stay where they were. Platforms gain coverage; the store economics absorb the drop. No quarterly release breaks this out. The only way to see it is to compare orders per store per day before and after an opening in the same polygon, which requires store-level tracking rather than city-level averages.
This is why a coverage number and a density number can move in opposite directions and both be true. Coverage counts doors. Density counts orders per door. A network can add 69 doors in a quarter and see density fall, and still call the quarter a success, because the promise being sold is speed of delivery, and speed is a radius problem dressed as a technology problem. Consumer demand is not the constraint; the cost of being close enough to serve it is.
Contribution margin is the only number that survives a new store
Contribution margin, expressed as a share of gross order value, is the closest thing to a store-level truth that a quarterly report offers. It nets out discounts, delivery cost and variable labour, and shows what each order contributes before fixed costs. Instamart's moved to negative 1.0 per cent from negative 2.6 per cent a year earlier, a 157 basis point improvement. Negative is still negative, but the direction matters: the marginal order is getting closer to paying for itself.
There is a structural change pending that should help. Shareholders approved capping foreign shareholding at 49.5 per cent so the entity qualifies as Indian owned and controlled, which would allow Instamart to hold its own inventory rather than operate on a marketplace basis. Management expects that to add roughly 80 basis points to contribution margin. If it lands, the network would move from negative territory toward zero without needing a single extra order.
| Metric | Q2 FY27 | Year earlier | What it tells you |
|---|---|---|---|
| Active dark stores | 1,240 across 133 cities | Not disclosed | Doors added, not density gained |
| Orders per store per day | 1,192 | Not disclosed | Utilisation of the average store |
| Contribution margin | -1.0% of GOV | -2.6% of GOV | Whether an extra order helps or hurts |
| Adjusted EBITDA loss | Rs 895 crore | Rs 849 crore | Cost of building ahead of demand |
| Net order value | Rs 6,594 crore | Rs 4,891 crore | Demand after discounts |
Two operators, two answers
The comparison that matters is not Instamart against its own past, it is Instamart against Blinkit. In the same September quarter, Blinkit operated 2,751 stores, grew net order value 79 per cent to Rs 20,940 crore, and posted adjusted EBITDA of Rs 184 crore, a swing from a Rs 156 crore loss a year earlier. It has said it expects more than 3,500 stores by March 2027.
Two networks, two very different states. Blinkit is adding stores and making money; Instamart is adding stores and losing more. The difference is not strategy on paper, since both run the same model of neighbourhood warehouses and short delivery windows. It is density and assortment mix at the store level. A network that fills its stores faster reaches contribution breakeven earlier, and every store added after that point improves the consolidated number instead of dragging on it.
What a brand should watch, not what the platform reports
For a consumer brand selling through these channels, store counts are a lagging indicator. They tell you where the network was built, not whether your product is pickable when someone orders it. The leading indicators are item-level fill rate by store, the share of your SKUs carried in the newest cohort of stores, and whether your listing survives the assortment cuts that come with each new opening.
A practical monitoring rule: pick the twenty stores that carry the highest share of your volume, track availability for your top ten SKUs in those stores weekly, and compare the newest cohort against the oldest. If availability in the new cohort sits well below the mature one, the network is expanding faster than its replenishment can follow, and your stockouts will concentrate exactly where the growth is.
What the next quarter has to show
Three numbers will settle the argument. The first is orders per store per day: if it climbs toward the 2,000 trigger while store count grows, the capacity is filling. The second is contribution margin, which should keep moving toward zero even before the ownership change lands. The third is the gap between mature and new stores, which nobody publishes but everyone can estimate from order-value totals divided by store count over time.
If all three move the right way, the quarter's loss was the cost of building a moat. If store count keeps rising while per-store orders stall, the network is buying coverage it cannot convert, and the next correction will be closures rather than openings. That distinction is worth more to a supplier than any headline about 133 cities.
FAQ
Does 1,192 orders per store per day mean the network is underused?
Not on its own. The figure is a mean across mature and new stores, and the disclosure does not break out the split. A network adding 69 stores in a quarter will always show a diluted average.
Why did the loss widen while the margin improved?
Because the network added fixed capacity faster than it added orders. Margin measures efficiency per rupee of order value; the absolute loss measures the size of the build.
What changes if Instamart can hold its own inventory?
Management expects roughly 80 basis points of contribution margin improvement from the ownership change. It affects margin, not demand, so it does not fix a store that cannot fill its capacity.
How should a brand use store counts in planning?
As a coverage signal only. For supply planning, use item-level availability in the stores that carry most of your volume, and compare the newest cohort against mature ones.
Is the 133-city footprint going to expand?
The footprint held at 133 cities while the store count rose, which points to deepening rather than widening. That shifts competition to density inside existing polygons, where a new store takes volume from a neighbour rather than from a new town.
