Editorial Article

Where China Put Its 2025 Subsidy Decided What Its Market Size Looks Like

Updated November 19, 2026

A subsidy has an address. In China’s 2025 quick commerce price war it was addressed to the order rather than to the delivery fee, and that single choice is still legible in the size of the market two years later.

The distinction sounds like accounting pedantry. It is not. Where a platform puts its money determines what its reported revenue looks like, what its market share looks like, and what the entire category’s size looks like — without changing a single customer’s experience or a single yuan of underlying economics.

Two ways to spend the same money

Imagine a platform willing to spend RMB 8 to acquire an order. It can spend it on the delivery fee, or on the goods.

Spend it on the delivery fee and the customer pays RMB 30 for the goods and nothing to have them delivered. The order value recorded is RMB 30 either way, because the delivery fee was never inside order value to begin with. The subsidy shows up in the contribution line as a cost of delivery, and the goods revenue is untouched.

Spend it on the goods — a coupon, a discounted line item, a first-order price — and the customer pays RMB 22 for the same basket. Now the order value recorded is RMB 22, and the market that gets measured is RMB 8 smaller. Same customer, same goods, same delivery, same subsidy. A different market size.

Why this matters more in China than anywhere else

In most markets the wedge is at least partly visible, because someone discloses both sides of it. India’s Swiggy Instamart reported gross order value of ₹7,881 crore and net order value of ₹5,674 crore for the same three months of Q4 FY26 — a ₹2,206 crore gap, 28% of the headline, in a single quarter. Anyone comparing an Indian platform’s gross number to another’s net number is measuring discount intensity and calling it share.

China offers no such luxury. No platform in the Chinese market publishes both a gross and a net figure, so the subsidy wedge has to be inferred from the direction of the basket rather than read off a disclosure. That is precisely why the market’s 1.22% annual basket decline should be read as a statement about accounting and competitive behaviour as much as about customers.

Where the money goes Effect on order value Effect on measured market size
Delivery fee waiver or reduction None. The fee sits outside order value. None. The subsidy is absorbed in the contribution line.
Item-level discount or coupon Reduces it by the full amount of the subsidy. Reduces it by the full amount of the subsidy.
Membership or subscription credit Depends on whether it is applied at checkout or amortised. Depends on the same, and rarely disclosed.
Free delivery threshold Raises it, by pushing baskets above the threshold. Raises it.

What Pheonix does about it, and what it costs

This research sizes China’s instant retail market on a net-of-subsidy base, from RMB 971.4 billion in 2025 to RMB 2,698 billion in 2033 at 13.62%, on a perimeter narrower than the official Ministry of Commerce research series. That choice is defensible — discount intensity is a competitive variable and should be visible separately rather than counted as demand — but it is a choice, and it has consequences.

The first consequence is that this number will read lower than any gross-basis estimate of the same market, and lower by an amount that is not a rounding error. The second is that it will diverge from a gross figure progressively across eight years, because the subsidy wedge is largest exactly in the years when platforms are buying share, and China’s price war is not over. The third is that the perimeter difference compounds alongside it: a whole-channel estimate includes operators that this build does not.

None of that makes either number wrong. It makes them answers to different questions. The useful discipline for anyone holding two estimates side by side is to ask where the subsidy sits and who is inside the count, in that order, before asking which growth rate is right.

The tell

There is one diagnostic worth carrying into any conversation about Chinese instant retail. A market whose order value falls while its order count rises is telling you that the growth is being paid for at the item level. A market whose order value and order count rise together is telling you that it is growing on demand. China’s 15.02% order growth against a 1.22% basket decline is unambiguously the first kind.

That does not make the growth less real. It makes it more expensive per yuan of revenue recorded, and it means the recorded market will keep understating the commerce actually being conducted for as long as the price war lasts.

Gross and net frames, the perimeter definition and the reconciliation against the official series are set out in the China Instant Retail Market report. Indian gross and net disclosures are drawn from Swiggy Limited’s Q4 FY26 shareholders’ letter.

Sources

The subsidy-placement argument is Pheonix analysis. Neither Meituan nor JD.com nor Alibaba discloses instant retail revenue net of subsidy, so the Chinese gross-to-net wedge cannot be measured the way Instamart’s can. That gap is the subject of the article, not a limitation of the sourcing.

Frequently Asked Questions

What is the difference between a gross and a net market-size figure?

Gross order value counts what the customer was charged before platform-funded discounts; net order value counts what was actually paid. The gap between them is the discount line, and in Swiggy Instamart's Q4 FY26 it was ₹2,206 crore — 28% of the gross figure — inside three months. Comparing one platform's gross number with another's net number measures discount intensity, not market share.

Why does where a subsidy sits change China's market size?

Because a coupon or an item discount reduces the amount recorded as order value, while a delivery-fee waiver does not — the fee sits outside order value entirely. Two platforms spending identically can therefore report different market sizes, and the one discounting goods will always look smaller for as long as it keeps discounting.

Do Chinese platforms disclose both gross and net order value?

No. No platform in the Chinese market publishes both a gross and a net figure, so the wedge has to be inferred from the direction of the basket rather than read from a disclosure. India's Swiggy Instamart does publish both, which is why the Indian wedge is measurable and the Chinese one is estimated.

Which base does Pheonix Research use for China?

A net-of-subsidy base on a perimeter narrower than the official Ministry of Commerce research series: RMB 971.4 billion in 2025 to RMB 2,698 billion in 2033 at 13.62%. That figure will read lower than a gross-basis estimate by an amount that grows across the forecast, because the subsidy wedge is largest in the years when platforms are buying share.