Editorial Article
Quick Commerce Takes 1.69 Per Cent of the Grocery Bill as Advertising. It Will Take 2.46 Per Cent.
Quick commerce platforms took USD 3.44 billion of advertising out of USD 203.6 billion of groceries in 2025 — 1.69% of the bill. By 2033 they will take 2.46%, or USD 14.70 billion of a USD 598.0 billion market.
The take rate moves only 77 basis points across eight years, which sounds like a business standing still. It is not. The advertising pool compounds at 19.92% while the groceries underneath compound at 14.42%, and that 5.5-point gap compounds into a profit pool roughly four times its current size without the take rate having to do much at all.
The arithmetic worth holding onto: the pool grows because the commerce under it grows, and then a bit more because it takes a slightly larger share of each order. The “bit more” compounds, and across eight years it is the difference between a pool that merely keeps pace with the groceries and one that outruns them.
Measured per order, eight cents
Advertising revenue was USD 0.0838 per order in 2025 across 41,000 million orders, reaching USD 0.1246 across 118,000 million orders in 2033. Orders grow 14.13% a year, revenue per order grows 5.07%, and the two compound to the 19.92% headline.
| 2025 | 2033 | Change |
|---|---|---|
| Advertising revenue USD 3.4371bn | USD 14.6986bn | 19.92% a year |
| Order value USD 203.6bn | USD 598.0bn | 14.42% a year |
| Take rate 1.69% | 2.46% | +77 basis points |
| Advertising per order USD 0.0838 | USD 0.1246 | 5.07% a year |
Eight cents does not sound like a profit pool until it is set against what the delivery of that order earns. Adjusted EBITDA in quick commerce is routinely a low single-digit share of order value, and at several operators it is negative. At 1.69% of the bill, the advertising line is not incremental to the business — on many orders it is larger than the contribution from delivering the goods.
Why the take rate is capped
The constraint on retail media inside a dark store is not advertiser demand. It is shelf space, in the digital sense: the number of sponsored placements a catalogue of 3,000 to 25,000 lines can carry before it stops being a shop and starts being an advert.
Push the sponsored share of a search results page past a certain point and customers stop trusting the ranking, which reduces the value of the ranking to the advertiser that paid for it. That ceiling is why the take rate rises 77 basis points over eight years rather than doubling, and why the growth in the pool comes mostly from the commerce rather than from the interception rate. Any forecast of retail media that shows the take rate tripling is a forecast of a worse shopping experience, not a better advertising business.
The most attributable inventory in retail
What quick commerce has that no other retail media format has is proximity. The session is seconds long, the catalogue is closed, the customer is at the point of purchase and the goods arrive in minutes. A supplier buying a placement can see the order it generated before the customer has finished unpacking.
That is why the format sustains a take rate at all against far larger retail media businesses with vastly more traffic. It is not selling impressions. It is selling a decision, at the moment the decision is being made, with the receipt attached.
Sizing is built from disclosed platform advertising revenue across the operating markets and applied to Pheonix Research’s instant retail order-value base. The take-rate bridge and the denominator reconciliation are in the Quick Commerce Retail Media Market report.
Sources
Both take rates in this article are disclosed advertising revenue divided by a disclosed transaction base — nothing is estimated at operator level:
- Instacart (Maplebear Inc.), Form 10-K for fiscal year 2025 — advertising and other revenue of USD 1,065 million against gross transaction value of USD 37,224 million, total revenue and adjusted EBITDA.
- Zepto Limited, UDRHP-I filed with SEBI, June 2026 — advertising revenue of INR 1,635.7 crore, advertising and promotional spend of INR 1,389.1 crore, operating revenue of INR 22,623.5 crore and net order value.
- Rohlik Group, audited annual report for the year to 30 April 2026 — advertising, trade marketing and retail media revenue against total revenue. Held by Pheonix Research; not published at a stable public URL, so named rather than linked.
The global take rate of 1.69% rising to 2.46% is a Pheonix model. Where no operator discloses a base — most of the rest-of-world block — the regional take rate is modelled and labelled as modelled in the report.
Frequently Asked Questions
How fast is the quick commerce advertising take rate rising?
It rises from 1.69% in 2025 to 2.46% in 2033 — 77 basis points across eight years — while the advertising pool grows from USD 3.44 billion to USD 14.70 billion at 19.92% a year. Most of the growth comes from the commerce underneath rather than from a higher interception rate.
Why is the take rate capped rather than compounding faster?
Digital shelf space. A catalogue of 3,000 to 25,000 lines can only carry so many sponsored placements before customers stop trusting the ranking, which destroys the value of the placement to the advertiser who paid for it. That ceiling is why the take rate rises 77 basis points rather than doubling.
What does advertising earn per quick commerce order?
USD 0.0838 per order in 2025, reaching USD 0.1246 in 2033 — eight cents rising to twelve and a half. It looks small until it is set against adjusted EBITDA, which in quick commerce is routinely a low single-digit share of order value and at several operators is negative. On many orders the advertising line is larger than the contribution from delivering the goods.
Why does quick commerce sustain a take rate at all against much larger retail media?
Because the placements sit inside a session that lasts seconds, in a closed catalogue, at the point of purchase, for goods arriving in minutes. A supplier can see the order a placement generated before the goods are unpacked. The format is not selling impressions; it is selling a decision with the receipt attached, which is why it sustains a take rate against retail media businesses with far more traffic.
