China Burned RMB 100 Billion on Food Delivery. The Bill Is Finally Arriving.
One year after China's tech giants ignited the most expensive food delivery subsidy war in the industry's history, the battlefield has gone quiet — leaving behind a trail of hemorrhaging balance sheets, 16 million surplus delivery riders, and a merchant class that collectively lost money while order volumes climbed.
The ceasefire is unmistakable. By early 2026, major platforms had begun winding down mass subsidy campaigns that had collectively consumed nearly RMB 100 billion (approximately US$13.9 billion) over the prior 12 months. The strategic retreat marks a watershed moment for China's instant retail sector—an industry analysts projected would surpass RMB 1 trillion (US$138.9 billion) in gross merchandise value in 2026—and raises fundamental questions about whether platform-driven, subsidy-fueled growth can ever translate into sustainable unit economics.
Market participants are now confronting a structural reckoning. Investors who had priced in winner-take-all dynamics are instead facing a three-way standoff where all combatants have been materially weakened, and the primary beneficiaries of the spending spree appear to have been neither platforms nor merchants, but coupon-harvesting consumers with zero platform loyalty.
Financials Expose How Deeply the War Burned Capital
The numbers from the most recent reporting periods are unambiguous in their severity.
JD.com posted group revenue of RMB 315.7 billion (US$43.8 billion) for the latest quarter, up 4.9% year-on-year. Net profit attributable to shareholders fell to RMB 5.1 billion (US$708 million), down 53.2% from RMB 10.9 billion (US$1.51 billion) a year earlier.
The damage was concentrated in JD's New Businesses segment—which includes JD Takeout, JD Property, Jingxi, and overseas operations—where operating losses widened sharply from RMB 1.3 billion (US$181 million) to RMB 10.3 billion (US$1.43 billion), implying an operating loss margin of 164.9%. In practical terms, JD was spending RMB 2.65 for every RMB 1 of new-business revenue generated.
Alibaba fared no better in its domestic commerce operations. For fiscal year 2026, the China commerce segment's adjusted EBITA reached RMB 107.5 billion (US$14.9 billion), a 44% year-on-year decline — a stark reversal for a segment that had long been the group's primary cash engine. The deterioration reflects both direct subsidy outlays and the margin compression that results when platforms compete on price rather than differentiation.
Meituan, the incumbent market leader, has not been immune either, though its established logistics density and merchant network provide a structural buffer that newer entrants lack.
Rider Oversupply Signals How Distorted Labor Markets Have Become
Perhaps the most vivid indicator of the war's collateral damage is the delivery rider imbalance. Industry estimates cited in research circulating in mid-2026 suggest the sector requires approximately 4 million active riders to service organic demand, yet the subsidy-driven order surge attracted upward of 20 million registered riders into the ecosystem — meaning more than five riders are now competing for every available order. The resulting income compression for gig workers is significant and largely invisible in platform-level financials.
This oversupply dynamic also illustrates the reflexive nature of subsidy-driven growth: platforms inflated order volumes artificially, which attracted labor supply, which inflated operational cost structures, which in turn required even larger subsidies to maintain competitiveness. The cycle was self-reinforcing on the way up and is now equally self-reinforcing on the way down.
Merchant Data Reveals the Subsidy Illusion Hiding Inside Revenue Growth
Academic and consulting research adds granular texture to what the macro financials suggest.
A study by Professor Zhang Jun's research group at Fudan University, based on transaction data from more than 40,000 restaurant merchants, found that following the intensification of platform subsidies after July 5, 2025, average daily total order volume (dine-in plus delivery combined) rose 7% — while average daily net receipts fell approximately 4% and total merchant profit margins declined 8.9%. Revenue went up; profitability went down. The subsidy was accruing to consumers and platforms, not operators.
A separate survey by Lixin Consulting of more than 2,000 food and beverage merchants painted an even starker picture: approximately 70% reported declining revenues, 80% reported falling net profit, and more than one-third saw net profit margins compress by more than 30 percentage points. For the majority of independent operators, participation in the delivery ecosystem during the war period was net value-destructive.
The structural explanation is straightforward. Average fulfillment costs per delivery order run RMB 6–8 (US$0.83–1.11), representing more than 15% of the average order value. For orders below RMB 30 (US$4.17)—the dominant price band in China's instant retail market—merchant margins are effectively wiped out once fulfillment expenses, platform commissions, and product costs are accounted for.
Scale Players Capture Disproportionate Gains, Widening Structural Divide
The war did produce winners, but they were concentrated among franchised chains with industrialized supply chains. Brands such as Luckin Coffee and Mixue Ice Cream & Tea — with raw material costs of under RMB 5 per cup against selling prices of RMB 15–20 — could absorb platform commissions of RMB 4–5 and still retain meaningful unit margins. Their standardized production processes, centralized procurement, and scale-driven cost structures made them structurally suited to high-volume, low-margin delivery economics.
Independent merchants had no equivalent buffer. The industry axiom that emerged from the war: of every 10 instant retail operators, roughly one generates meaningful profit, three break even, and six are effectively subsidizing platform growth with their own labor.
Even the instant retail vertical beyond food delivery struggled. Jiubian, an alcohol-focused instant retail operator, reported revenue of RMB 1 billion (US$138.9 million) in 2025, down 39.7% year-on-year, while posting a net loss of RMB 161 million (US$22.4 million). Front-warehouse operators who entered the market expecting to capitalize on the instant retail boom found themselves trapped between rising rents, high labor costs, and inventory write-downs.
AI Deployment Optimizes Platform Extraction, Not Merchant Survival
One of the war's more counterintuitive outcomes involves artificial intelligence. All three major platforms deployed AI-driven dispatch, inventory replenishment, and customer service systems throughout 2025, and efficiency metrics — delivery times, inventory turnover, support costs — measurably improved. But the productivity gains did not flow to merchants or riders. Instead, they lowered platform operating costs while commission structures remained unchanged, effectively allowing platforms to extract more value per transaction without sharing efficiency dividends downstream.
This dynamic — where AI investment concentrates returns at the platform layer while atomized suppliers bear unchanged or increasing cost burdens — may become a template for how algorithmic optimization reshapes power dynamics in China's platform economy more broadly, with implications for sectors ranging from automotive retail to consumer electronics.
Subsidy Withdrawal Forces Industry Toward Unit Economics Discipline
The pivot to profitability that began in early 2026 is not voluntary restraint — it is financial necessity. With investor patience eroding and cash reserves under pressure, the platforms have little choice but to rationalize subsidy spending. The strategic question now is whether any player can translate temporary volume gains into durable market share before the subsidy withdrawal triggers user churn.
Historical analogues are not encouraging. China's group-buying wars of the early 2010s saw hundreds of platforms collapse into one. The shared bicycle boom consumed tens of billions of renminbi before consolidating into a duopoly atop a graveyard of hardware. In both cases, subsidy-acquired users evaporated when economics normalized.
The instant retail sector's structural path forward requires either higher average order values to support fulfillment economics, or fundamental reductions in per-order delivery costs — likely through route density optimization, autonomous delivery pilots, or consolidation of order flows. Neither solution is achievable at scale in the near term.
For investors holding positions in JD.com, Alibaba, or Meituan, the key variable to monitor is not gross merchandise volume but the trajectory of segment-level operating margins in new business lines through the second half of 2026. The war is over. The reconstruction — and the bill — is just beginning.
Related Coverage:
China's Tech Giants Bleed $14 Billion in Food-Delivery War With No Winner