Starbucks sold control of China — and changed what its $37 billion in revenue means

Chipotle owns every store, so its revenue is its scale. Domino's (opens in new tab) owns almost none, so its $20 billion system collapses to a $5 billion company. Starbucks sits between them — and is the most revealing of the three, because it shows that the mix of owned versus licensed isn't fixed. It's a dial a company can turn. In November 2025, Starbucks turned it hard: it sold a controlling 60% stake in its entire China business — nearly 8,000 stores — to the investment firm Boyu Capital, in a deal valuing the China unit at around $4 billion. Overnight, the single largest block of stores Starbucks operated became something it merely licenses.

That is why Starbucks' headline number deserves a second look. The company reported $37.18 billion in total net revenue for FY2025 — but that figure is not one thing. It is the sum of three very different streams, and the China deal is about to change the recipe.

One number, three different things

Break the $37.18 billion apart as Starbucks itself reports it:

• Company-operated stores — about $30.7 billion. Sales rung up at the roughly half of Starbucks' locations the company owns and runs. This is operational scale in the purest sense, the only piece comparable to a Chipotle.
• Licensed stores — about $4.35 billion. Crucially, this is not the sales of licensed stores. It's the income Starbucks earns from licensing — royalties plus the coffee, cups, and equipment it sells to licensees who run their own stores in airports, grocery chains, and overseas markets. The licensed stores' actual register sales never appear in Starbucks' revenue at all.
• Other — about $2.1 billion. Packaged coffee and ready-to-drink products sold through grocery (the global coffee-aisle business), foodservice, and royalties from the partnership with Nestlé. A consumer-packaged-goods business with nothing to do with a café.

So a reader who treats $37.18 billion as "how big Starbucks' stores are" is wrong three times over. Roughly half of Starbucks' stores are owned but they generate close to 87% of revenue; the licensed half contributes only a thin royalty-and-supply line; and a chunk of the total is supermarket packaged goods. The most common error this invites is dividing total revenue by total store count to get an "average unit volume." Don't: the right per-store figure uses only company-operated revenue over company-operated stores — about $1.43 million — and even that flatters the picture, because it excludes the lower-volume licensed base.

The China reclassification — caliber in real time

Now layer the Boyu deal on top, because it is one of the cleanest live demonstrations of why scale labels matter.

Before the sale, Starbucks China's nearly 8,000 stores were company-operated: every cup sold there flowed into that $30.7 billion company-operated line. After the deal closes — expected around the second quarter of fiscal 2026 — Boyu controls 60% of a joint venture that owns and runs those stores, while Starbucks keeps a 40% minority interest and licenses its brand and intellectual property to the venture. In accounting terms, Starbucks will deconsolidate China: those store sales leave its revenue entirely, replaced by equity-method income on the 40% stake plus a licensing stream — exactly the structure of the "licensed" bucket above.

[!guardrail] FY2026 revenue will be basis-broken — a reclassification, not a collapse The consequence is the kind of thing FoodBud exists to flag: Starbucks' reported revenue will fall by several billion dollars, and the company will not have shrunk at all. The same coffee, the same 8,000 stores, the same brand — simply moved from the "owned" column to the "licensed" column. Anyone comparing FY2026 revenue to FY2025 without knowing this will misread a reclassification as a collapse. It is the same lesson FoodBud draws from Yum! Brands and the separately listed Yum China (opens in new tab): the brand's footprint and the parent's revenue live on different layers, and an ownership change can move billions between them overnight.

Why Starbucks let go of China

Starbucks didn't sell control of its second-largest market from a position of strength. It sold because the China coffee war turned against it. Luckin Coffee, the homegrown rival that overtook Starbucks as China's largest coffee chain back in 2023, has built past 26,000 stores in Asia on a model Starbucks can't easily match: a small Americano runs roughly 10 yuan (about $1.40) at Luckin versus around 30 yuan (about $4.21) at Starbucks, sold through an app-first, discount-driven loyalty machine. Luckin even opened its first U.S. stores in New York in 2025. Against a faster, cheaper local operator, Starbucks chose a local partner with the capital and market savvy to compete — taking $4 billion off the table now, keeping 40% of the upside, and continuing to collect brand royalties. Whether that proves shrewd or a retreat, it is unambiguously a bet that Starbucks is better as China's licensor than its operator.

"Back to Starbucks": the turnaround under an ex-Chipotle chief

The China deal is one front in a broader fix-it campaign run by a familiar name. Brian Niccol — the CEO who built Chipotle's digital turnaround before leaving in 2024 (see our Chipotle deep-dive (opens in new tab)) — took over Starbucks in September 2024 and launched a strategy he calls "Back to Starbucks": less transactional app-and-go, more of the coffeehouse experience that built the brand. The diagnosis was that Starbucks had optimized itself into long mobile-order queues, an overcomplicated menu, and stores that had stopped feeling like the "third place."

The fixes are operational and concrete. A new "Green Apron" service model reprioritizes throughput and face-to-face hospitality; an early pilot of 650 stores outperformed the rest of the fleet by about 200 basis points of comparable sales. Starbucks is adding 25,000 seats back to its U.S. company-operated cafés, simplifying the menu, and relaunched its loyalty program in three tiers (Green, Gold, Reserve). Stores are now measured on a short internal scorecard — sales, throughput, staffing, customer satisfaction, food safety.

After a soft FY2025 — North America comparable sales fell 2% on a 4% drop in transactions — the early returns are real. By the first half of fiscal 2026, U.S. comparable sales had swung to their first gains in two years, rising about 7% in the most recent quarter; global comps turned positive; and even China returned to growth as the handover loomed. Niccol's own framing is that the company is "ahead of schedule." The market, valuing the equity near $118 billion, is paying for the turnaround to keep working.

Unit economics and the model

The owned base is where Starbucks makes its money, and its quality shows: company-operated revenue runs about $1.43 million per store, with a North America segment operating margin around 11.5% — healthy for a business carrying real estate, baristas, and the cost of being a "third place" rather than a grab-and-go counter. The strategic logic of the mixed model is that Starbucks owns the markets where the brand experience and economics justify the capital (the U.S. above all) and licenses the rest — airports, supermarkets, and increasingly, now, China. The dial between owned and licensed is, in effect, a capital-allocation decision dressed up as an operating-model choice.

The caliber takeaway

Starbucks is the capstone of this three-part series because it proves the central point in motion. Chipotle's revenue equals its scale because it owns everything; Domino's revenue is a fraction of its system because it owns almost nothing; Starbucks sits in between — and just slid further toward the Domino's end by handing China to a licensee. Do not read its $37.18 billion as homogeneous store sales: it blends company-operated sales, licensing income, and packaged goods. Do not divide total revenue by total stores for an average unit volume. Do not compare FY2026 revenue to FY2025 without adjusting for the China deconsolidation — a multi-billion-dollar drop that reflects ownership, not size. And do not reach for the ~$118 billion market capitalization as a measure of scale; it prices the equity, nothing more.

"How big is Starbucks?" has, as of this year, an even more layered answer than before — which is precisely the point. The number didn't get bigger or smaller. The company changed what it counts.

Starbucks Corporation (SBUX) — the data card

Caliber notes. Scale basis = total net revenue, which blends three different natures (company-operated store sales, licensing/royalty income, and CPG/foodservice) — not directly comparable to a pure company-operated chain's revenue (Chipotle) or to a franchisor's system sales (Domino's). Forward change: on close of the Boyu deal, Starbucks deconsolidates China — reported revenue will drop by several billion with no change in the brand's footprint; treat any FY2025→FY2026 revenue comparison as basis-broken until restated, and thereafter treat China as a licensee-overlap (same structure as Yum/Yum China). Do not divide total revenue by total store count for AUV. USD reporter — no FX conversion. Market cap fenced from all scale comparisons. Licensed-store system sales are not disclosed in Starbucks revenue and are not summed here.

Sources. Starbucks Q4 & full-year FY2025 results (29 Oct 2025; fiscal year ended 28 Sep 2025); Starbucks FY2026 first-half results and "Back to Starbucks" / Green Apron disclosures (2026); Starbucks–Boyu Capital China joint-venture announcement (Nov 2025); Luckin Coffee competitive coverage (2025); FoodBud locked operator record (markguog/foodservice-listed-operators, export 2026-06-14). Cross-references: Chipotle (C9), Domino's (C10), Yum! Brands / Yum China (C2).