- Quick-commerce platforms quote a category commission, but the number that runs your P&L is the all-in take rate: commission + fulfillment + gateway + platform fees + ads + returns reserve. For most brands that lands between 30% and 38% of MRP.
- The same take-rate stack hits a ₹120 SKU and a ₹900 SKU very differently, because fulfillment is a flat fee. Below a price floor, no amount of volume makes the channel profitable.
- Compute your maximum sustainable ad load — contribution margin before ads, minus the profit you require — before you set a single bid. Most q-commerce ad "waste" is really a budget that was impossible at that margin.
Quick commerce is the fastest-growing shelf in Indian retail — the sector cleared roughly ₹11,000 crore of GMV in January 2026 alone, and every D2C founder has felt the pull. The dashboards are seductive: orders land within hours of going live, conversion runs far above your website, and GMV climbs week on week. Then the first monthly settlement arrives, and the number in the bank is barely half the GMV the dashboard celebrated. Nobody lied. You just hadn't done the stack math yet.
The take-rate stack, line by line
| Layer | Typical range | Charged on |
|---|---|---|
| Category commission | 12–22% (staples lowest, beauty & nutrition highest) | MRP — usually the list price, not your discounted price |
| Fulfillment / logistics fee | ₹8–25 per order | Flat per order — brutal on low-ticket SKUs |
| Payment gateway | ~2% prepaid | Order value |
| Platform / cataloguing fees | Fixed monthly, often waived at volume | — |
| Returns & damage reserve | 0.5–2% held back | Weekly payouts, released later |
| Ads (the voluntary layer) | 8–14% of GMV for most actively-advertising brands | Whatever you bid |
Two things make the headline commission misleading. First, everything else: Morgan Stanley pegged Blinkit's blended take rate near 18% for a recent quarter, and reporting in Business Standard put the all-in share retained by platforms — once storage, warehousing and delivery charges are counted — at 30–35% of selling price. Second, the base: commission is typically computed on MRP, so when you fund a 10% discount, the platform's cut doesn't shrink with your price. Your discount comes entirely out of your side of the split.
Worked example: the same stack at three price points
Assume a mid-range category: 16% commission, ₹18 fulfillment, 2% gateway, 10% of GMV on ads, COGS at 40% of MRP, sold at full MRP. Watch what the flat fulfillment fee does as the ticket size drops.
| ₹120 SKU | ₹450 SKU | ₹900 SKU | |
|---|---|---|---|
| Commission (16%) | −₹19 | −₹72 | −₹144 |
| Fulfillment (flat) | −₹18 (15% of MRP!) | −₹18 (4%) | −₹18 (2%) |
| Gateway (2%) | −₹2 | −₹9 | −₹18 |
| Ads (10% of GMV) | −₹12 | −₹45 | −₹90 |
| Net realisation | ₹69 (57%) | ₹306 (68%) | ₹630 (70%) |
| COGS (40%) | −₹48 | −₹180 | −₹360 |
| Contribution per unit | ₹21 (17%) | ₹126 (28%) | ₹270 (30%) |
The ₹120 SKU gives up 15% of MRP to fulfillment alone — the ₹900 SKU gives up 2%. Same category, same commission letter, wildly different businesses. This is why "we're doing great on Blinkit" and "we lose money on every quick-commerce order" are both common, and both true, sometimes inside the same brand's catalogue. The channel has a price floor, and it's arithmetic, not opinion: below roughly ₹150–200 MRP, most categories cannot carry a flat fulfillment fee and an ad load and COGS and still keep anything.
Your maximum ad load is a formula
Take the ₹450 SKU: after commission, fulfillment, gateway and COGS — but before ads — contribution is ~38% of MRP. If you require 15% operating profit from the channel, your ad budget can consume at most 23% of GMV. If you're a scaling brand happy to run the channel at breakeven for share, the ceiling is 38%. Either way it's a computed number, and it converts directly into the only ROAS target that means anything on these platforms: spend ₹1, and the GMV it generates must carry the whole stack. We'll do that math in the companion piece on quick-commerce ROAS — the short version is that a "profitable" campaign at 4× reported ROAS can be underwater once the take rate is priced in.
The working-capital tax nobody budgets
- Settlement lag: payouts typically run T+7 or longer, with tighter terms only at volume. You buy inventory today; the platform pays you next week for what sold.
- Returns reserve: a slice of each payout is held back against damages and disputes, released weeks later. Small per week, permanent as a float.
- Fill-rate pressure: dark stores penalise stockouts — in ranking and sometimes in fees — so you carry inventory across dozens of micro-warehouses. That's working capital multiplied by geography.
Where this leaves the channel
None of this is an argument against quick commerce. High-frequency, mid-ticket, impulse-friendly categories can clear real contribution there, and shelf presence has a discovery value the math above doesn't capture. The argument is narrower: treat the platform settlement report — not the GMV dashboard — as the source of truth, and price every growth decision at the contribution line. The brands that get burned aren't the ones with bad products; they're the ones who scaled a take-rate stack they never itemised.