China’s Delivery War Is Over: What Instant Retail Now Means for Imported Food Brands
Updated
Three Chinese platforms spent around 150 billion RMB in twelve months trying to buy each other’s delivery customers. In late March 2026 the regulator told them to stop. The subsidies are winding down, the discounts your Chinese consumer got used to are going away, and the channel that emerged from the wreckage is now one of the fastest routes to a Chinese shopper for an imported food or drink brand.
I am Jon Wang, founder of SEO Agency China. We work with food and drink brands that sell into China without a local office. Almost none of them had instant retail in their channel plan two years ago. Most of them should have it now. Here is what actually happened, and what it means if you sell a physical product people eat or drink.
The five numbers that matter
- Around 150 billion RMB burned by Meituan, Alibaba and JD.com between Q2 2025 and Q1 2026, roughly 21 billion USD at 7.1 RMB to the dollar (Huxiu, 2026).
- Split of that spend: Alibaba around 70 billion RMB, Meituan around 40 billion, JD.com around 35 billion (same source).
- Peak volume passed 200 million orders a day, up from 80 to 90 million before the war started (same source).
- Q1 2026 daily orders: Meituan around 65 million, Taobao Shangou around 50 million, JD.com around 9 million. Roughly a 5:4:1 split.
- Late March 2026: the State Administration for Market Regulation signalled the war should end, and all three pulled back.
The questions brands ask me about instant retail
Is this just food delivery with a new name?
No, and the difference decides whether your product belongs there. Section two.
My product is not urgent. Why would anyone order it in 30 minutes?
That was the right question in 2024. It is the wrong one now, and section three explains why.
Do I need a Chinese entity for this?
Practically, you need stock inside China and somebody who owns it. That is the real gate, not the paperwork. Section five.
The subsidies are ending. Did I miss the window?
You missed the cheap window. You did not miss the channel. Those are different things, and the business reminder at the end explains the difference in money.
Will my distributor block this?
Often yes, because it touches their price. This is the conversation to have before you sign, not after.
What actually happened, in order
Alibaba decided it could not let Meituan own the daily habit of ordering things to your door. JD.com decided it could not sit out. Meituan decided it would defend. From Q2 2025 the three of them paid Chinese consumers to change behaviour, and Chinese consumers obliged.
| Platform | Spend Q2 2025 to Q1 2026 | Daily orders, Q1 2026 | Where it goes next |
|---|---|---|---|
| Alibaba (Taobao Shangou) | ~70 billion RMB | ~50 million | Pushing instant retail from emergencies into routine restocking, tied to the Taobao account |
| Meituan | ~40 billion RMB | ~65 million | Defending the high-basket user, moving into instant retail and own-operated stock |
| JD.com | ~35 billion RMB | ~9 million | Supply chain first, volume second |
Then the arithmetic caught up. Nobody makes money paying people to order a bubble tea. Margins collapsed, the platforms started trimming subsidies on their own, and in late March 2026 the market regulator made it official by circulating an Economic Daily piece arguing the delivery war should end. The most aggressive phase closed.
What did not close is the habit. 150 billion RMB bought a permanent change in how Chinese consumers expect to get groceries, snacks, drinks and household goods. That behaviour does not unwind when the coupon stops.
Why this became a channel and not a promotion
The shift that matters for a food brand is not the delivery speed. It is what people order.
Instant retail started as an emergency service. You ran out of something, you paid a premium, it arrived. Alibaba’s move was to drag it from emergency into routine, using the Taobao account and its shopping history to make the 30-minute order a normal weekly restock rather than a rescue. That changes the basket. An emergency basket is one item. A restock basket has eight, and it has room for a brand the shopper is trying for the first time.
It also changes who your competition is. On Tmall you compete against every imported brand in your category, worldwide, on a search results page. On instant retail you compete against what is physically in a warehouse within a few kilometres of the shopper. That is a much shorter list.
For a small imported brand this is the most favourable competitive geometry in Chinese e-commerce right now. Getting into one city’s local stock is a smaller fight than ranking nationally.
Which imported products work here, and which do not
Not every product belongs in a 30-minute basket. The honest filter is simple: does someone want it today, and is the price low enough that they will not research it first?
| Category | Fit | Why |
|---|---|---|
| Imported snacks, biscuits, chocolate | Strong | Impulse, low ticket, no research needed, high repeat |
| Beer, RTD, soft drinks, water | Strong | Occasion-driven, heavy to carry, classic instant retail behaviour |
| Dairy, butter, cheese, chilled | Good, with cold chain | Restock behaviour, but the logistics cost decides the margin |
| Coffee and tea | Good | Daily habit, and the occasion is often “now” |
| Premium spirits, fine wine | Weak | High ticket, researched, gifted. The buyer wants provenance, not speed |
| Supplements, infant nutrition | Weak to mixed | Trust-led purchase, heavy regulation, buyer compares before ordering |
If you are in the bottom two rows, the honest advice is to leave instant retail alone and put the money into search visibility and the private domain instead. Being available in 30 minutes does not help a product somebody spends two weeks deciding to buy.
A second opinion from the marketplace side
Instant retail is an operations problem before it is a marketing one. I asked John Parker, E-Commerce Director at E-Commerce China Agency, who has tracked the subsidy war from inside the platforms, what he tells brands now:
The brands that got hurt are the ones that treated the subsidy price as their real price. They let the platform discount the product for a year, the consumer anchored on it, and now the subsidy is gone and the brand looks expensive overnight. Protect your price floor from day one, even when the platform is paying for the discount. You will still be selling this product when the promotion budget has moved on.
John Parker, E-Commerce Director, E-Commerce China Agency
That lines up with what we saw from the search side. When a Chinese shopper sees the same imported product at three prices in one week, the first thing they do is search the brand name to work out which price is real. If your brand has no Chinese content answering that, the answer comes from a forum thread.
How to enter without a Chinese company
Instant retail runs on stock that is already in China. Cross-border parcels from a bonded warehouse do not move in 30 minutes. So the entry question is who holds your inventory.
Through a distributor who is already on the platforms
The fastest route, and the one most small brands should take. Your distributor already stocks local warehouses and already runs Meituan and Taobao Shangou storefronts. You are adding SKUs to an existing operation. The trade-off is that you do not control the price, which is exactly the risk John Parker describes above. Put a price floor in the contract. We cover how to qualify these partners in our guide to finding a reliable Chinese distributor.
Through a local agent running your store
More control, more cost, and you carry the inventory risk. It makes sense once one city works and you want to protect the brand rather than just move volume. This is the same operating model as a marketplace flagship, which is why brands usually do it after Tmall rather than before.
The part that decides whether either works
Instant retail has almost no space to tell a story. A listing is a photo, a name, a price and a delivery time. Nobody reads a brand history in a 30-minute basket. So the discovery has to happen somewhere else, before the order.
In practice that means the shopper has already met your brand on Xiaohongshu or Douyin, or looked you up. That is the work we do: making sure that when someone types your brand into Baidu, or asks DeepSeek or Doubao whether an imported brand is any good, there is something in Chinese to find. Our Baidu SEO and AI search work exists for that moment. The social seeding that creates the demand in the first place runs through Xiaohongshu, where imported food and drink still gets discovered.
Business reminder: what this channel costs and what it returns
Take the platform drama out and look at it as a channel decision.
What it costs. Three things, and only one of them is a marketing budget. First, stock sitting in China, paid for before it sells, in a market where you cannot easily pull it back. Second, a platform commission and a delivery cost that together eat a bigger share of a low-ticket item than most European brands expect. Third, the price you can never quietly raise again, because instant retail prices are visible, compared and screenshotted. The subsidy era trained an entire consumer base on a discounted number. Whoever sells at that number next is paying for it out of their own margin.
What it returns. Repeat purchase, which is the thing imported food brands in China are chronically bad at. A Tmall Global order is often a one-off gift or a trial. An instant retail order is a restock. If your product is genuinely good and genuinely repeatable, this channel compounds in a way that cross-border does not. The second benefit is speed of proof: you find out whether Chinese consumers actually rebuy your product in weeks, in one city, for a small stock commitment. That is a cheaper market test than a national launch, and the answer is more honest.
One number to hold on to. The platforms spent around 150 billion RMB to move Chinese consumers from 80 million daily orders to over 200 million. That behaviour change is paid for and permanent. You are inheriting it for the cost of a pallet and a listing, which is the cheapest piece of consumer conditioning a foreign brand will ever be handed in China. The mistake is assuming the discount is part of the inheritance. It is not.
The trap. Going in because the volume looks good, without a price floor and without demand already existing. Instant retail does not create demand. It converts demand that already exists, faster. A brand nobody has heard of, sitting in a local warehouse, with no Chinese content behind it, sells nothing at all, and the unsold stock is now your problem in a country where you have no entity.
Our recommendation. Do not treat this as a national channel. Pick one city, one distributor, a small stock commitment and a hard price floor, and spend more on the Chinese-language visibility that makes people want the product than on the listing itself. If it rebuys in that city, expand. If it does not, you learned something important for the cost of a pallet.
The question worth arguing about
One view says the subsidy war destroyed price discipline in Chinese food retail for years, and a foreign premium brand should stay out until the market re-anchors.
The other says the habit is permanent, the platforms now need real products rather than discounts, and a brand that enters while the platforms are courting supply gets terms it will never get again.
I lean towards the second for snacks and drinks, and towards the first for anything premium. Where do you land, and what would change your mind? Tell me.
Sources
- Huxiu, 2026, on the cost of the delivery war: the ~150 billion RMB total, the split between Alibaba, Meituan and JD.com, the peak of over 200 million daily orders, and the Q1 2026 order shares.
- TMTPost, 2026, on the regulatory tightening and the shift from subsidy-driven growth to competition for the consumption entry point.
- NetEase, 2026 instant retail industry report, on how the category is being restructured after the subsidy phase.
- Our own channel guide: China, the number one food importer, for the wider picture on what gets imported and through which route.
Jon Wang founded SEO Agency China. He spends most of his week with imported food and drink brands deciding which Chinese channel to try next, usually with less budget than the decision deserves. Connect with Jon on LinkedIn.
Working with us on this
We are fifteen people in Jing’an, Shanghai. On instant retail we do not run your logistics. We do the part that decides whether the listing sells: Chinese content that makes your product findable and credible before the order, on Baidu, on the Chinese AI assistants, and on Xiaohongshu.
One case from our own work, anonymised at the client’s request. A European snack brand was already listed through a distributor on two instant retail platforms in Shanghai and selling almost nothing. The listing was fine. The problem was that searching the brand name in Chinese returned nothing except the listing itself, so nobody had a reason to pick it over a brand they recognised. We built the Chinese-language basics, product facts, origin, and a handful of seeded reviews with real users. Reorder rate in that city moved enough for the distributor to extend to a second city the following quarter. No new spend on the platform itself.
If you are weighing this channel, tell us the product and the city you are considering and we will tell you honestly whether it is worth the stock.