From Starbucks to Cotti: Luckin's Rivals and Its 'Anti-Fragile' Logic

This is an English adaptation of a FoodBud historical article originally published on February 26, 2025.

Luckin Coffee's recovery from its 2020 fraud scandal to China's largest coffee chain is a case study in turnaround governance and an efficiency-led operating model. (Note: the original is a long, narrative piece; this adaptation keeps the facts and operator lessons and trims rhetorical color.)

From near-collapse to revival

In April 2020 Luckin disclosed about RMB 2.2 billion in fabricated transactions. The stock fell ~75–80% in a day, the SEC opened an investigation, and Luckin became the first US-listed Chinese company to settle with US regulators (a USD 180M settlement). It closed nearly 3,000 of ~4,500 stores; employee attrition topped 40%; franchisees protested. Four years later it had revenue of RMB 24.9 billion and net profit of RMB 3.03 billion (2023), surpassing Starbucks China, with 22,000+ stores.

The governance reset

Centurium Capital injected USD 700M for a controlling ~40% stake (with IDG and Qatar's sovereign fund following), liquidating the founders' shares. Centurium's "three no's" — no team change, no operational interference, no performance bet — paired with two demands: full data transparency to shareholders, and equity incentives covering ~300 core staff. Management retention exceeded 90%; supplier terms were renegotiated (payables cut from 45 to 30 days); loss-making franchisees got subsidies and a no-forced-termination pledge (renewal ~70%). By end-2020 cash flow turned positive. A cultural reset followed — a defined mission, an internal taste-test committee, a store-manager partner profit-share, and flattened information access.

The efficiency moat

Luckin rebuilt unit economics around a pick-up model:

• Format: ~90% small pick-up stores (20–50 sqm vs. Starbucks' ~150 sqm third-place), cutting rent share from ~30% to ~10%.
• Labor: full-automatic machines plus SOPs cut staff from ~8 to ~3 per store; productivity ~130 cups/person/day (vs. Starbucks ~80).
• Cost: ~RMB 6.5 per cup all-in (vs. Starbucks ~15% higher on beans), supporting a 35% gross margin even at RMB 9.9 pricing; AI ordering cut spoilage from 8% to 2%.
• By end-2021: ~400 cups/store/day and gross margin recovering from -33.1% to 65.8%.

Digital and product engines reinforced it: ~120M private-domain (WeChat) users with layered retention (repurchase to 54%); referral coupons cutting CAC sharply; blockbuster SKUs (the Coconut Latte peaked at 2M+ cups/day, ~30–40% of sales); 100+ new SKUs a year with a ~35% hit rate. Supply-chain positioning — multi-year Brazilian bean contracts, locked Southeast Asian coconut capacity, and roasting capacity scaling toward 150,000 tons by 2025 — anchors cost control.

Competition and global probe

Against Cotti's store-count chase, Luckin leans on cost (about RMB 1.2/cup lower bean cost; 10-point lower rent share) and AI-driven low spoilage. Internationally it is testing direct operation in Singapore (47 stores) before franchising (Malaysia first), with a supply-chain-first approach. Management frames "9.9 yuan" not as a price war but as the output of a lower cost structure. Forward views (China coffee growth "another 10 years," overseas requiring structural cost change) are from early 2025; the 2024 Q1 ~RMB 100M loss that reversed to a ~RMB 1.5 billion Q3 profit shows the model's volatility and flexibility.