Quick commerce is the category most investors in India have an opinion on, and the one where the published numbers are most useful for testing that opinion. I have not worked in it. I have run a business whose economics depend on the same variable, the density of demand around a fixed asset, so this is how I read it.
What the numbers say
Eternal, Zomato's parent, reported that Blinkit ended the March 2026 quarter with 2,243 dark stores and adjusted EBITDA of ₹37 crore on net order value of ₹14,386 crore, after reaching adjusted EBITDA profit for the first time in the preceding quarter. The shareholder letter reported about 27.2 million monthly transacting users, an average order value of ₹525 and roughly 274 million orders in the quarter.
Zepto, its nearest rival, reported FY25 turnover of ₹11,110 crore, about 2.5 times the ₹4,454 crore of FY24, with a net loss of ₹3,367 crore, up from about ₹1,215 crore.
The numbers underneath the numbers
| Derived figure | Working | Result |
|---|---|---|
| Blinkit adjusted EBITDA margin | ₹37 crore ÷ ₹14,386 crore | About 0.26% of order value |
| Orders per dark store per day | 274 million ÷ 2,243 stores ÷ about 90 days | Roughly 1,350 |
| Adjusted EBITDA per order | ₹37 crore ÷ 274 million orders | Roughly ₹1.35 |
A business making about ₹1.35 per order has almost no room for error. A small rise in delivery cost per order, a slight fall in order value or a few hundred fewer orders per store per day would put it back into loss. The profitability is real, and it is a thin line that holds only at very high density.
Density is the business
A dark store is a fixed asset with rent, staff and inventory that cost the same whether it serves 300 orders a day or 1,500. Each extra order within delivery range spreads that cost further and makes the rider's route shorter. This is the same logic as utilisation in an electric cab fleet, where a car's fixed cost only pays back if it is busy. At about 1,350 orders per store per day, Blinkit's network is dense enough to work. A new store in a thin neighbourhood starts far below that and loses money until it fills up.
That is why I would be careful about quick commerce in smaller cities, where the same store would see a fraction of the orders. I set out that caution in an operator's thesis on building beyond the metros.
Who else the ten-minute promise changes
For consumer brands, quick commerce is a new shelf that sells the top few products in each category and pays attention to availability hour by hour. For neighbourhood stores, it competes directly on the small, urgent basket that used to be theirs. For the platforms, the next margin is expected to come from higher-value categories, advertising sold to brands, and private labels, each of which raises the value of an order without adding delivery cost.
What I would ask a quick commerce business
- Orders per store per day, for stores open more than a year against new stores.
- How many months a new store takes to reach breakeven.
- Delivery cost per order, and how it moves with order density.
- Contribution per order with discounts and before advertising income.
- The share of stores that are profitable today.
