This article used to be about instant Nescafé and a handful of Starbucks stores selling a Western lifestyle to young professionals. That version of the Chinese coffee market is gone. What replaced it is a price war that gutted margins across the whole sector, a Starbucks that no longer owns a majority of its own China business, and a domestic supply chain in Yunnan that finally produces beans good enough for Starbucks itself to use.
Olivier Verot is the founder and CEO of GMA, based in Shanghai since 2012. He has watched three waves of foreign F&B brands try to read this market, and the coffee sector has changed faster than almost any other category his clients work in.
The price war that ate the industry’s margins
Luckin Coffee (瑞幸) started it, Cotti Coffee (库迪) kept it going, and the 9.9 yuan cup became the price every chain in the country had to answer to. The logic was simple: undercut Starbucks so hard that trying coffee stops feeling like a Western indulgence and starts feeling like buying a bottle of water. It worked on volume. It did not work on profit.
Luckin’s own numbers show the cost. In Q3 2025 the company’s net revenue rose 50.2% year-on-year to 15.29 billion yuan, but operating profit grew only 12.9%, and the operating margin slid from 15.5% a year earlier to 11.6%, according to 36Kr’s analysis of the widening gap between Starbucks China and Luckin. Growth at that speed should widen margins, not shrink them. It didn’t, because every new store still has to compete on a 9.9 to 13.9 yuan cup.
Cotti is feeling it harder. The chain announced a target of 50,000 stores for 2025, pushed a low-cost “Cotti Express” kiosk format to hit it, and still closed 722 locations in a single 90-day stretch through June 2026, about 4.5% of its entire footprint. Luckin has started quietly narrowing its own 9.9 yuan promotions to a handful of base drinks and letting most of the menu drift back up to 10.9 to 13.9 yuan. Nobody wants to say the price war is over. Everybody is trying to walk away from it without losing face.
We’ve written separately about the brands that are refusing to play this game at all: China’s Tea and Coffee Shakeout looks at why the chains still standing five years from now will be the ones that built loyalty instead of a discount habit.
Starbucks China: what actually happened
On November 3, 2025, Starbucks announced it was selling a 60% stake in its China retail business to Boyu Capital, a Chinese private equity firm, for $4 billion. Starbucks keeps the remaining 40% and becomes, functionally, a licensor: it owns the brand, collects royalties, and no longer controls day-to-day operations of its roughly 8,000 China stores. Starbucks itself put the total value of the arrangement, sale proceeds plus retained stake plus a decade or more of licensing income, at over $13 billion. The deal is expected to close in Starbucks’ fiscal Q2 2026.
Read the target the new joint venture set for itself: grow from 8,000 stores to 20,000. That number only makes sense if you accept that Starbucks can no longer win on brand prestige alone in China. It needs a local partner who can move at Luckin’s speed. Twenty-six years after entering the market, Starbucks gave up sole control of it. That’s not a company retreating. It’s a company admitting its old playbook stopped working.
We covered the mechanics of how Luckin got here in detail elsewhere: How Luckin Beat Starbucks in China: 5 Lessons for F&B Brands. That article is worth reading if you’re deciding whether to fight for premium positioning or scale positioning. This one is about where the whole market stands today.
The market in numbers
| Metric | Luckin Coffee | Cotti Coffee | Starbucks China |
|---|---|---|---|
| Stores (2025) | ~29,200 globally (99.6% in China) | ~10,000 by early 2025, chasing 50,000; closed 722 in 90 days by June 2026 | ~8,011 stores, net add of only 415 for the full 2025 fiscal year |
| Ownership | Independently listed, Chinese-founded | Independently owned, Chinese-founded | 60% Boyu Capital, 40% Starbucks (from Q2 FY2026) |
| Typical price | 9.9 to 13.9 yuan | 3.9 to 9.9 yuan, delivery-subsidy heavy | 25 to 35 yuan |
| 2025-2026 trajectory | Revenue up 50%+, margins compressing | Store closures rising despite expansion target | Ceded majority control, targeting 20,000 stores long-term |
The market underneath these three players is not shrinking. iiMedia Research puts China’s total coffee industry at just over 1 trillion yuan in 2025, up from 623.5 billion yuan in 2023. What changed is who captures that growth, and at what margin.
Specialty coffee is rising in parallel, not instead
While Luckin and Cotti fight over the 10 yuan cup, a second, quieter market has been building in Shanghai, Beijing, Chengdu, and Shenzhen: independent roasters, single-origin cafés, and chains like Manner and Seesaw that charge Starbucks prices or more and still fill tables. These two markets barely compete with each other. The 9.9 yuan drinker is buying caffeine and habit. The specialty customer is buying a bean’s origin story, a barista’s technique, and a place to be seen.
iiMedia’s own consumer data backs this up: close to a third of Chinese coffee drinkers now drink two to three cups a week, and the frequent-drinker segment keeps growing even as average spend per cup on the cheap end falls. Volume and premiumization are climbing at the same time. That’s unusual, and it’s the opening a foreign specialty brand actually has, one that a brand trying to out-discount Luckin never will.
The channel most foreign brands still miss: selling online, not in stores
Almost every foreign brand that asks us about China coffee starts from the same assumption: they need a store, or at least a corner in someone else’s. They don’t. Premium beans, capsules, and instant coffee sell well on Tmall, JD, and increasingly Douyin, and cross-border e-commerce lets a brand ship into China before it has a local entity, a licensed kitchen, or a single square meter of retail space. We laid out how the compliance side of that works in China CBEC Law 2026: What Will Change for Foreign Brands, and where the traffic itself is moving in Why Brands Are Leaving Tmall for Douyin.
Margot, who runs a small French specialty roastery, is a good example of what that shift looks like in practice. She tried the traditional route first: a pop-up café in a Shanghai mall, six months, one location. Rent alone ran close to 40,000 yuan a month, and she was competing for foot traffic against a Luckin selling drinks at a third of her price two doors down. The pop-up broke even on a good month and lost money on a normal one.
What changed her numbers was closing the café and moving the same beans, roasted and bagged in France, onto a cross-border storefront on Tmall Global, with a bonded-warehouse listing that avoided the general-trade registration her volume didn’t yet justify. No rent, no staff, no competing on the price of a cup she never serves. Her store did just under 1,800 orders in its first full quarter, at three to four times what a supermarket instant coffee costs per serving, sold to people who were never going to choose her over a 9.9 yuan Luckin cup anyway. The lesson wasn’t that retail doesn’t work in China. It’s that she was fighting Luckin’s fight instead of her own.
Yunnan is no longer a footnote
China doesn’t just drink coffee now, it grows a serious amount of it. Yunnan produces over 98% of the country’s coffee beans, and the 2025-2026 harvest is on track to be one of the largest on record: overall output up 30 to 40%, with some producing regions up 50%, on roughly 1.2 million mu of plantation and around 140,000 tons of green beans.
The bigger change is quality, not volume. Specialty-grade Yunnan beans now sell at roughly a 60% premium over ordinary commodity-grade beans, growers are expanding plantings of Gesha, Typica, and Bourbon varieties instead of relying on generic Catimor, and Yunnan coffee has moved from a discount ingredient into Starbucks’ own China supply chain, reportedly in every espresso shot the chain pulls in mainland stores and in several of its single-origin offerings. A decade ago, “Yunnan coffee” meant cheap and forgettable. It doesn’t anymore.
For any brand sourcing beans, blending them, or simply wanting a defensible “made with Chinese-grown coffee” story on packaging, Yunnan is now a credible answer, not a compromise. Brands exploring direct sourcing relationships often start at China’s major food trade fairs, where growers, importers, and roasters meet in person; we’ve mapped the main ones in China’s Food Industry: 5 Key Food Fairs You Can’t Miss.
FAQ
Is Starbucks pulling out of China?
No. Starbucks sold 60% of its China retail business to Boyu Capital in a deal announced in November 2025 and expected to close by fiscal Q2 2026, but it kept 40% and stays on as brand owner and licensor. It’s a loss of operational control, not an exit. The new joint venture is targeting 20,000 stores over time, more than double the current count, which suggests Starbucks expects to keep growing in China, just under someone else’s day-to-day management.
Should a new coffee brand compete on price against Luckin and Cotti?
Almost never, unless you can match their store density and their delivery-platform subsidies, which a new foreign entrant usually can’t. Luckin’s own operating margin fell as it chased volume at 9.9 to 13.9 yuan, and Cotti is closing hundreds of stores while still expanding. Compete on origin, quality, or a channel they don’t touch, like premium retail or online, instead of matching their cup price.
Can we sell coffee in China without opening a single store?
Yes, and for most foreign brands it’s the more sensible starting point. Roasted beans, capsules, and instant coffee move well through Tmall Global, JD Worldwide, and Douyin, using cross-border e-commerce rules that let you ship into China before setting up a local entity. It won’t build the same brand visibility as a physical café, but it removes rent, staffing, and direct price competition with Luckin from the equation entirely.
Is Yunnan coffee actually good, or just cheap?
It’s improved faster than most people outside the industry realize. Specialty-grade Yunnan beans now command around a 60% premium over commodity grade, growers are planting more Gesha, Typica, and Bourbon, and Starbucks itself sources Yunnan beans for its China espresso base and some single-origin lines. It’s not going to replace Ethiopian or Colombian beans on a specialty menu overnight, but it’s no longer a quality compromise, just a cheaper one.
GMA advises food and beverage brands on China market entry, from cross-border e-commerce compliance to retail and distributor negotiations. If you’re deciding between a physical footprint and an online-first launch for a coffee, tea, or specialty food brand, that’s a conversation we have with clients every month. Reach out through our contact page to talk through what fits your volume and budget.
