Across forty interviews in July 2026, mature tier two dark stores ran a store level contribution margin of four to seven percent, roughly three points below the figure reported centrally. The difference is rider incentives, local marketing and fresh wastage, and it closes as order density passes about nine hundred a day.
- Surat7.5%
- Jaipur6.3%
- Lucknow5.2%
- Indore4.1%
- Nagpur3.0%
- Coimbatore2.1%
- Kochi0.9%
- Bhubaneswar−0.5%
What is quick commerce in a tier two market?
Quick commerce is grocery and convenience retail delivered from a small local warehouse, usually within ten to thirty minutes. The warehouse is called a dark store: it holds two to four thousand items and serves customers only through an app, never in person.
In a metro that description is enough. In a tier two city it is not, because two things change at once. The catchment a single store can cover is larger, since traffic is lighter and a rider covers more ground in the same fifteen minutes. And the order density inside that catchment is lower, often by half. One store therefore serves more area with fewer orders, which is the whole of the economic problem in a sentence.
The eight cities in this study were Jaipur, Indore, Lucknow, Coimbatore, Nagpur, Bhubaneswar, Surat and Kochi. All eight had at least two national operators live for twelve months or more at the time of the interviews.
Why reported and store level margins diverge
Every city lead we spoke to could quote a contribution margin for their stores. In every case it was lower than the number they understood to be reported upward. Nobody described this as a discrepancy, and nobody suggested anything was being concealed. They described a boundary question: which costs belong to the store.
Three costs sat on that boundary in almost every conversation.
- Rider incentive during peak. Base rider cost is charged to the store. The surge paid between seven and ten in the evening is frequently held centrally as a demand generation cost.
- Local marketing. Launch offers, first order discounts and neighbourhood activation are typically a city or national line, though the store is the only thing they move.
- Fresh wastage. Charged to the store in five of the eight cities and to a regional shrinkage pool in the other three.
None of these is an accounting error. Each is a presentation choice, and each is defensible. The reason it matters for anyone underwriting the asset is that the choice differs by operator, so two contribution margins quoted in two data rooms are frequently not the same measurement.
If you ask me what the store makes, I will tell you after riders and after wastage, because that is what I am judged on. The number that goes up does not have the peak incentive in it. Both are true.
What the numbers look like across eight cities
The table below sets the two figures side by side. Reported is the contribution margin the city lead understood to be circulating centrally. Store level is the same measure once the three costs above were charged to the store. Both are expressed as a percentage of net order value.
| City | Orders / day per store |
AOV (₹) |
Reported margin |
Store level margin |
Gap (points) |
|---|---|---|---|---|---|
| Surat | 1,180 | 486 | 8.9% | 7.5% | 1.4 |
| Jaipur | 1,040 | 512 | 8.1% | 6.3% | 1.8 |
| Lucknow | 960 | 468 | 7.4% | 5.2% | 2.2 |
| Indore | 880 | 455 | 6.8% | 4.1% | 2.7 |
| Nagpur | 820 | 441 | 6.2% | 3.0% | 3.2 |
| Coimbatore | 760 | 498 | 5.9% | 2.1% | 3.8 |
| Kochi | 690 | 523 | 5.1% | 0.9% | 4.2 |
| Bhubaneswar | 610 | 430 | 4.4% | −0.5% | 4.9 |
Read down the density column and the pattern is hard to miss. Average order value moves within a narrow band across all eight cities, and Kochi has the second highest basket of the set while sitting seventh on store level margin. Density explains the ranking. Basket size does not.
Where the cost actually sits
Charging the three boundary costs back to the store produces a consistent breakdown. Rider incentive accounted for an average of 2.1 points, local marketing for 0.7, and fresh wastage for 0.5. The wastage figure is the widest of the three, running from 0.6 percent of net order value in cities with a same day replenishment run to 3.4 percent in cities without one.
Rent behaved differently from expectation. Dark store rent per square foot in these cities ran at roughly a third of the metro equivalent, but the stores were larger, because a wider catchment needs deeper stock. Rent as a share of net order value therefore landed close to metro levels rather than far below them.
What changes as a city matures
Operators consistently placed the inflection at roughly nine hundred orders a day per store. Below it, the store is carrying a cost base sized for a catchment it is not yet filling. Above it, the same riders and the same shelf serve more orders and the boundary costs stop mattering, because they shrink as a share of a larger base.
Two of the eight cities had crossed that line during the twelve months before the interviews. In both, the reported and store level figures had converged to within two points, and in both the city lead described the store network as having stopped needing local marketing support to hold volume.
This is also why break even timing varied so widely. Fourteen to twenty months was the range described for stores opened since 2024, against eight to eleven months in the metros. The variable named was never basket size and never rent. It was how quickly the catchment reached density.
How we ran this
- Sample
- Forty semi structured interviews, 30 to 45 minutes each. Twenty two city leads, thirteen dark store managers, five regional operations heads.
- Coverage
- Eight cities: Jaipur, Indore, Lucknow, Coimbatore, Nagpur, Bhubaneswar, Surat, Kochi. Four national operators, none named at their request.
- Field window
- 2 to 24 July 2026. Every figure quoted refers to the June or July trading month unless the article says otherwise.
- Known limits
- All figures are self reported and none has been reconciled to audited accounts. Three of the eight cities are represented by a single operator, so the city medians there are effectively one company's experience. Compliance screening excluded four candidates who were still employed by a listed operator in a role with visibility of unpublished results.
What this means if you are underwriting the asset
The practical consequence is that a contribution margin in a data room is not comparable to one in another data room until you know which costs sit inside it. Three questions settle it quickly, and all three can be answered by an operator who has run a store rather than by the seller.
- Is peak rider incentive charged to the store or held centrally?
- Which line carries first order discounting, and for how long after a launch?
- Is fresh wastage a store cost or a regional pool, and has that changed in the last two years?
Those are the questions we now put into scoping for this sector. If you are working on a live process, expert calls with two or three city leads will answer all three inside a week, and our consumer and retail bench covers all eight of these cities.