Luckin Coffee entered 2026 under pressure from a structural cost shift that its rapid expansion model was never designed to absorb. The company's fourth-quarter 2025 results, released before U.S. market open on Feb. 26, laid bare a fundamental tension: delivery-driven revenue growth is coming at a price that is increasingly difficult to justify.

The headline numbers told a story of divergence. Total revenue rose 32.9% year-over-year to RMB 12.78 billion yuan (approximately US$1.76 billion), but Non-GAAP operating profit fell 13% to RMB 960 million yuan — a combination that missed market expectations and underscored a structural margin problem rather than a cyclical one. The culprit was not competition in the traditional sense, but a delivery subsidy war that has fundamentally disrupted the low-cost fulfillment model on which Luckin built its profitability.

The market had been pricing in a faster recovery. Investors expected that as platform subsidies from third-party delivery aggregators gradually wound down, Luckin's self-pickup ratio would rebound, pulling delivery cost ratios back toward historical norms. That normalization is happening — but far more slowly than anticipated, and the Q4 data suggests the structural drag may persist well into 2026.

The Delivery Cost Trap

The core of Luckin's margin problem is a number that stands in stark contrast to the company's original business logic: delivery costs as a percentage of revenue hit 12.8% in Q4 2025, up sharply from 8.7% in the same period a year earlier. In absolute terms, delivery expenses surged 94% year-over-year to RMB 1.63 billion yuan. Under normal operating conditions — before the delivery subsidy wars intensified — that ratio sat comfortably between 7% and 9%.

Luckin's entire unit economics model was architected around self-pickup. By concentrating stores in high-footfall locations and encouraging customers to collect orders directly, the company minimized last-mile fulfillment costs and achieved operating margins that were exceptional by quick-service beverage standards. The delivery wars, ignited by aggressive subsidies from major food delivery platforms, effectively forced Luckin into a high-cost fulfillment channel it had deliberately avoided.

The subsidy environment in Q4 2025 was not as extreme as the peak seen in July and August, but it had not normalized either. Based on channel checks, Q4 subsidy intensity was broadly comparable to May-June 2025 levels — meaning the structural distortion persisted through the entire second half of the year. Platform commissions and delivery fees are directly compressing Luckin's margins, pulling a self-pickup-centric model into the economics of a delivery-dependent one.

Pricing Strategy: Defending Margins Without Losing Customers

Faced with rising fulfillment costs, Luckin pursued a two-pronged pricing strategy in Q4 that reflects the difficulty of raising prices in China's hyper-competitive coffee market.

On the promotional side, the company significantly reduced the scope of its RMB 9.9 yuan pricing. Products qualifying for the 9.9 price point were cut from 8-10 SKUs in Q3 to just two — standard Americano and latte — while the frequency of "buy-two-get-one" promotions and large-denomination coupons was materially reduced. Discount channels on third-party platforms, including group-buy and flash-sale formats, also saw their discount rates compressed.

On the product side, Luckin avoided direct price increases on core SKUs such as its Coconut Latte and Velvet Latte, but introduced value-added options — oversized cups, extra espresso shots — that effectively increased average ticket size without triggering consumer backlash. New product launches in Q4, including a Minions collaboration and a winter wellness latte series, carried price points 30%-50% above the base product range.

The strategy is analytically coherent: maintain the "affordable" brand positioning to defend against ultra-low-cost competitors such as Xingfu Coffee, while extracting incremental margin from customers willing to trade up. However, the Q4 same-store sales growth (SSSG) of just 1.2% year-over-year — a significant deceleration from Q3 — suggests that consumer price sensitivity is higher than management anticipated. Volume growth disappointed even as per-cup revenue edged up, indicating that the pricing power Luckin is attempting to assert remains constrained.

User Growth Slows as Subsidy-Driven Retention Proves Shallow

Monthly active paying users reached 98 million in Q4 2025, up 27% year-over-year but decelerating meaningfully compared to Q2 and Q3 growth rates. The pattern reveals a structural weakness in the delivery-driven user acquisition model: customers attracted by platform subsidies exhibit low retention rates once those subsidies are withdrawn.

The delivery wars effectively brought a large cohort of highly price-sensitive users into Luckin's ecosystem. These users, acquired at significant promotional cost, have shown limited willingness to repurchase at normalized price points — a dynamic that is now visibly dragging on monthly active user growth. The conversion rate from delivery-acquired users to Luckin's proprietary membership ecosystem appears to be underwhelming, limiting the long-term value of what was, in effect, subsidized customer acquisition.

On the positive side, Luckin's paid membership penetration continued to rise, reaching 28%-30% of users in Q4 — up approximately 3 percentage points from Q3. Paid members spend 30%-40% more per transaction than non-members, providing a partial offset to the volume shortfall. The company's Lucky AI 1.0 personalization system, launched in Q3, is credited with improving targeted promotional efficiency and supporting paid membership conversion.

Store Expansion: Deliberate Slowdown, Selective Acceleration

Luckin added a net 1,834 stores in Q4 2025, bringing its total network to 31,048 locations — crossing the 30,000-store milestone while still meeting its full-year target of 8,000 net new openings. However, the pace of expansion was notably slower than in Q3, a shift the company appears to have made deliberately.

The logic is straightforward: when delivery costs erode store-level operating margins, the return on capital from opening new stores deteriorates. Prioritizing unit economics over raw store count is a rational response to the current margin environment. Luckin simultaneously increased subsidies to franchised stores — particularly those with below-average cup volumes — to help franchisees navigate the seasonally weak first quarter of 2026. This reflects both the company's recognition that it expanded aggressively into lower-tier markets in 2025 and its continued conviction that franchised store density in those markets remains strategically important.

The company's direct-operated stores, which now exceed 20,000 locations concentrated in Tier 1 and Tier 2 cities, account for more than 80% of profits. Franchised stores, surpassing 10,000 locations and focused on lower-tier markets, contribute approximately 20% of revenue. Franchise revenue grew 39.3% year-over-year in Q4 — faster than the 31.2% growth from company-operated stores — driven largely by raw material sales to franchisees, which account for nearly 70% of franchise segment revenue.

International Expansion: Singapore as Proof of Concept

While domestic operations navigate the delivery cost cycle, Luckin's Southeast Asian expansion is generating early validation. In Singapore, where the company operates 81 fully company-owned stores as of Q4 2025, the business has reached overall profitability — a meaningful milestone that Luckin is positioning as its international benchmark.

The Singapore model replicates the domestic dual-format approach — fast-pickup and dine-in — with stores prioritized in central business districts, university campuses, and transit hubs. The 100% company-owned structure preserves brand control and operational standards.

In Malaysia, Luckin has adopted a franchise model, signing a 10-year exclusive licensing agreement with GASB, a subsidiary of a local conglomerate with substantial commercial real estate assets. GASB manages full in-country operations while Luckin provides brand licensing, digital systems, supply chain support, and training. The 70 stores currently operating, concentrated in Kuala Lumpur and Penang, are structured as 100% fast-pickup locations. The model is designed as a template for further expansion into Indonesia and Thailand. North America remains at an early stage, with the company still refining its single-store unit economics against the dual headwinds of high rents and low brand recognition.

Outlook: Structural Disruption or Temporary Headwind?

The central analytical question for Luckin investors in 2026 is whether the delivery cost overhang represents a permanent structural shift or a temporary distortion that will self-correct as platform subsidies normalize.

The case for transience is grounded in the economics of delivery subsidization: platform subsidies are not indefinitely sustainable, and as they recede, the incentive for consumers to order via delivery — rather than self-pickup — diminishes. As self-pickup ratios recover, delivery cost ratios should trend back toward the 7%-9% normalized range. Regulatory scrutiny of delivery platform subsidy practices adds further pressure toward normalization.

The case for persistence rests on consumer behavior change. Once customers develop delivery habits, reverting to self-pickup is not automatic, even after subsidies disappear. If a meaningful portion of the delivery-acquired user base has structurally shifted its ordering behavior, Luckin's cost base may not fully normalize even as subsidy intensity fades.

The Q4 data does not definitively resolve this question, but the trajectory — subsidy intensity in Q4 comparable to May-June 2025 rather than reverting toward pre-war levels — suggests normalization is slower than the market assumed. With China's overall coffee penetration still rising and the competitive landscape requiring continued investment in store density and product innovation, Luckin's medium-term growth thesis remains intact. But the path back to peak margins runs through a delivery cost normalization that, based on current evidence, is likely to be measured in quarters rather than weeks.