1,240 dark stores at Instamart and 2,751 at Blinkit, as of 10 November 2026, and both numbers went to investors as evidence of progress rather than of capacity. Neither company published sales per outlet, the payback period on a site opened this year, or the share of outlets trading below the chain average. For a brand that plans distribution coverage in India, that gap decides whether a new city is a market or a line on a map （Livemint）.
What The Quarter Actually Disclosed
Blinkit added 308 net outlets in the quarter and told investors it expects to pass 3,500 by March 2027 and about 5,000 by March 2028, from 2,751 at the end of September, with a large share of the additions coming from smaller cities （Livemint）. Instamart added 69 net outlets to reach 1,240 across 133 cities, a network that reaches more cities than its rival's but holds fewer sites inside each one.
The economics moved the other way in the same release. Instamart's contribution margin fell to negative 1 per cent of gross order value from negative 0.2 per cent in the preceding quarter, while net order value rose 34.8 per cent to 6,594 crore rupees and the adjusted core loss widened to 895 crore from 849 crore. Blinkit reported net order value of 20,940 crore rupees, up 79.3 per cent, and an adjusted EBITDA profit of 184 crore, its sixth consecutive quarter of margin improvement.
| Measure | Instamart | Blinkit |
|---|---|---|
| Net outlets, end September 2026 | 1,240 | 2,751 |
| Net additions in the quarter | 69 | 308 |
| Cities served | 133 | Not broken out in the same detail |
| Net order value, year on year | +34.8 per cent | +79.3 per cent |
| Contribution line | -1 per cent of gross order value | Adjusted EBITDA +184 crore rupees |
Swiggy described the build differently on the same call, saying the network had moved to larger and more efficient sites holding nearly three times the SKUs of a year earlier, with the footprint reaching about 2 million square feet, 54 per cent more than a year before （Economic Times）. That is a capacity statement per site. It is not a statement about whether the site earns its rent, and in the first year of a new market the two are only loosely related.
Independent reading of the same numbers reaches the same limit. A review of the results notes that the disclosures do not establish how each location performs, how quickly individual stores reach viability, or what occupancy and operating costs are in different markets （Urban Acres）. A network can be expanding and still be unreadable at the level where a supplier has to make a decision, and that is the level where range reviews, promotional funding and shelf space are actually decided.
Why An Outlet Count Cannot Stand In For Coverage
Coverage, in the sense a supplier uses the word, is a listing decision measured store by store and week by week. An outlet count is a construction decision measured quarter by quarter. The two move at different speeds, which is why a network can grow by a quarter of its size and still show flat on-shelf availability in the new cities for two full quarters. Treating the count as coverage pushes too much stock into cities where picking is still stabilising and too little into cities where the format has already settled.
The trade area problem sits underneath that. A dark store serves a radius defined by delivery time rather than by footfall, so two sites three kilometres apart can chase the same households and compete for the same order. Nothing in a quarterly count says which of the two carries the volume, and nothing in it says whether a site opened in July is now trading at the chain average or at half of it. A supplier that maps its own distribution coverage onto the platform's outlet count inherits that blindness without noticing.
The store potential question
Store potential in this format is not a property of the building. It is a joint function of household density, the share of orders that arrive inside the promised window, and the share of those orders filled from the shelf without substitution. A site can sit in a dense trade area and still fail the third measure, and that failure appears in the topline as a demand problem rather than as a picking problem, which is why it survives so long in a category report.
The Three Numbers A Brand Should Ask For
The first is orders per outlet per day, split by city tier. A national average hides the fact that a mature city and a city entered this year differ by a factor no promotion closes. The second is the filled rate on the ten SKUs that carry the category, measured at outlet level: a chain-level number is the average of a good store and a bad one, and the average is what a planogram meeting usually gets. The third is the substitution rate, because a shopper who receives a different pack has been served by the platform and lost by the brand.
Ask for those three on the same list, for the same week, and the outlet count stops being the headline. A network that added 308 sites and a network that added 69 can then be compared on the only axis that matters to a supplier: how often the pack was on the shelf when the order arrived, in the cities where the growth is supposed to happen.
What the substitution rate hides
Substitution is usually reported as a service metric and it is really a distribution metric. When a pack is replaced, the platform keeps the basket and the brand loses the sale, so the failure never shows up in a category sales number. Across a network that is adding sites quickly, substitution clusters in the newest outlets, which is exactly where a range review is least likely to look and where a listing is most likely to be assumed rather than checked.
A Four Week Test For An Outlet List
None of this needs a new system. It needs a list, a fixed week and the willingness to order the same basket twice, which is a smaller commitment than a national availability report and arrives sooner than a quarterly review cycle. The list is the unit of evidence, and four weeks is enough time to see whether a filled rate is moving in the right direction.
- Ask the platform for the outlet list it will confirm in writing, with city and opening month for each site, and keep that file as the baseline for every later comparison.
- Split the list into sites older than twelve months and sites younger, then treat the two groups as two different networks rather than one.
- Order the ten category SKUs from five outlets in each group on the same weekday, and record for each line whether the pack arrived, was substituted or was cancelled.
- Repeat in week four against the same list, then report the filled rate rather than the order value, because value moves with promotion and the filled rate does not.
The traceability table matters more than the score it produces. When a filled rate drops between week one and week four, the record shows whether the drop came from a specific pack size, a specific outlet or the whole city, and that distinction is the difference between a conversation with a category buyer and a conversation with a supply chain team.
What Changes For The Holiday Window
The last six weeks of the year are when outlet additions and promotional calendars collide. A site that opened in September is being asked to hold a festive range it has never carried, at the moment when picking capacity is stretched by order volume. That is the window in which a national availability figure looks acceptable while the ten SKUs that carry the category are missing in the stores that matter, and it is the reason a supplier should fix its outlet list before the calendar does it.
FAQ
Does a larger outlet network mean better availability?
Not by itself. Availability depends on whether each outlet is replenished at the rate it sells, so a network can grow while the filled rate in the new sites stays below the chain average for several quarters.
Why ask for sales per outlet if platforms rarely publish it?
Because the answer, even a refusal, tells you whether the number exists. A platform that tracks it can rank sites; one that cannot is choosing sites on catchments alone, which is a different risk for a supplier funding a listing.
Is substitution always bad for the brand?
It is bad when the substitute is another brand and neutral when it is another pack of the same brand. That is why the substitution rate should be recorded at pack level rather than at order level.
How often should the outlet list be rebuilt?
Every quarter, and immediately after any expansion announcement. A list that is six months old describes a network that no longer exists in the cities where growth is happening.
