McDonald’s: Control the System Without Operating Every Restaurant
McDonald’s scale is not explained by franchising alone. The deeper design is a split between local execution and retained control rights: most restaurants are operated by franchisees, while McDonald’s preserves the common system, brand standards, sales-linked economics and, in conventional franchises, important real-estate rights. The model works only while local operators can invest, execute and earn enough to stay aligned with the system.
The Brief in One Sentence
McDonald’s shows that a company can decentralize most day-to-day operations without decentralizing every right that holds the economic system together.
Why It Matters
At June 30, 2026, McDonald’s reported 46,028 restaurants worldwide. 44,016 were franchised and 2,012 were company-operated. In other words, roughly 95% of the network was run by operators that were not McDonald’s employees.
That scale can be misunderstood. Franchising does not mean McDonald’s simply hands over a logo and collects a fee. The company describes multiple franchise structures, common standards, sales-linked revenues and, in conventional franchises, a lease-and-license architecture in which McDonald’s usually retains control of the underlying real estate and building when the typical 20-year term ends.
The useful question is therefore not “Does McDonald’s own the restaurants?” It is: which responsibilities are local, and which rights stay central?
Explain It Simply
Imagine a school with hundreds of cafeterias. Each cafeteria has a local manager who hires the team, serves customers and handles the daily problems. But the school keeps the name, the menu rules, the safety standards, the ordering system and sometimes the building itself.
The local manager has freedom to operate, but not freedom to turn the cafeteria into an entirely different business. That is closer to the logic of McDonald’s than the simple idea that “someone else owns the restaurant.”
Evidence Map
- Observed / year-end 2025: McDonald’s had 45,356 restaurants; approximately 95% were franchised.
- Observed / June 30, 2026: the System had 46,028 restaurants, including 44,016 franchised and 2,012 company-operated restaurants.
- Observed / franchise structure: franchised restaurants operate under conventional franchise, developmental license or affiliate structures; the capital and property arrangements differ.
- Observed / conventional franchise: the arrangement generally combines a lease and a license. Franchisees pay initial fees plus continuing rent and royalties based on sales.
- Observed / 2025 franchised revenue: $16.548 billion, including $10.442 billion of rents, $6.018 billion of royalties and $88 million of initial fees.
- Observed / property: net property and equipment under franchise arrangements was $22.8 billion at year-end 2025, including $7.1 billion of land.
- Observed / current digital layer: McDonald’s reported nearly 220 million 90-day active loyalty users across 70 loyalty markets at Q2 2026 quarter-end.
- Inference: the company’s architecture is better understood as a split between decentralized execution and retained control rights than as either “restaurants McDonald’s owns” or “real estate McDonald’s rents.”
- Uncertain: public filings do not provide a universal restaurant-level decomposition of franchisee returns, local investment payback, control value or the incremental contribution of each system standard.
The First Split: Who Operates?
The franchisee carries much of the daily operating burden: local labor, service, execution, occupancy-related costs and the responsibility to run the restaurant within the McDonald’s System. This makes the local operator economically consequential rather than a passive license holder.
McDonald’s, meanwhile, can grow the System without placing every restaurant’s workforce, daily operating decisions and full local execution directly inside the corporation.
That is decentralization — but only one layer of it.
The Second Split: What Stays Under Control?
In conventional franchising, McDonald’s generally combines the restaurant license with control of the site through ownership or a long-term lease. The franchise arrangement is usually about 20 years. When that term ends, McDonald’s says it retains control of the underlying real estate and building and can re-franchise the restaurant to the same operator, choose another operator or close the location.
The company also retains the common system: brand standards, operating requirements, products, technology interfaces and other rules that protect a recognizable customer experience.
This is the deeper architecture: local execution can move outward while selected control rights stay inward.
Why the Real-Estate Story Is Only Half True
Real estate is economically important. In 2025, rents represented $10.442 billion of McDonald’s $16.548 billion in franchised-restaurant revenue. The company also reported $22.8 billion of net property and equipment under franchise arrangements.
But calling McDonald’s simply a real-estate company removes the rest of the mechanism. A vacant restaurant site does not generate the same economics as a productive McDonald’s operated by an aligned franchisee under a strong brand and common system. Royalties are tied to sales. Rent economics depend on restaurant activity. Brand demand and operator execution remain essential.
The property strengthens control; it does not replace the restaurant business.
Not Every Franchise Has the Same Capital Model
The conventional-franchise structure is not the whole network. Developmental licensees generally provide the capital for their restaurant businesses, including real-estate interests, and McDonald’s generally does not invest restaurant capital in those businesses. McDonald’s instead receives sales-linked royalties and generally initial fees.
This distinction matters because the simple phrase “McDonald’s controls the real estate” is not universally true. The company adjusts the ownership structure by market, legal environment, available operators and capital conditions.
Franchisee Economics Are a Core System Dependency
A franchised model can make corporate revenue look more stable, but it does not make the operator’s economics irrelevant. McDonald’s filings explicitly identify franchisee financial health, cooperation and willingness or ability to reinvest as material to System performance.
This is logical. If local restaurants cannot earn enough to maintain equipment, hire well, renovate, adopt technology or meet standards, the center eventually inherits a weaker brand and a weaker revenue base.
The company can decentralize costs and execution. It cannot decentralize the consequences of widespread operator weakness.
Digital Makes the Shared System More Important
McDonald’s reported nearly 220 million 90-day active loyalty users across 70 loyalty markets at the end of Q2 2026. Digital ordering, loyalty, delivery and common technology create more interfaces that cut across individual restaurants.
That can strengthen the center because customer identity, offers, data, ordering and technology become more standardized. It can also create new tensions over technology cost, local economics, data use and how fast operators must invest.
Digital scale therefore does not eliminate the franchise relationship. It makes the quality of coordination more important.
What Most People Miss
The most transferable lesson is not “use franchises.” It is that organizational scale can be designed by splitting rights and responsibilities deliberately.
Some decisions benefit from local knowledge and owner-like attention. Others must remain common or the network stops behaving like one system. The hard work is deciding where that boundary sits — and updating it when technology, regulation, economics or customer behavior changes.
Where the Model Can Break
- Franchisee economics weaken. Operators may delay reinvestment, resist initiatives or exit if returns become unattractive.
- Control becomes too rigid. Central standards can protect consistency but can also slow local adaptation if the boundary is badly designed.
- Control becomes too loose. Food safety, labor practices, service failures or local misconduct can damage the common brand.
- Property commitments become burdensome. Long-lived sites and leases are valuable only when restaurant demand remains productive.
- Technology creates misalignment. Systemwide platforms can require capital and operating changes whose benefits and costs are not shared equally.
- Regulation changes the contract. Franchise, labor, property, data and competition rules can alter how control rights can be exercised.
Critical View — What Not to Copy
- Do not copy the franchise percentage. A high franchise share is not a strategy by itself.
- Do not copy rent extraction without operator economics. The network needs financially viable operators.
- Do not buy property merely because McDonald’s does. Property control works here inside a high-volume restaurant system and differs by franchise type.
- Do not centralize every standard. The lesson is to choose the boundary, not to maximize central control.
- Do not call the model asset-light without qualification. McDonald’s still carries substantial property and system investment.
Sidy’s Synthesis — Delegate the Work, Keep the Architecture
My synthesis is that McDonald’s is best read as a design of control boundaries.
The corporation does not need to own every local operating decision. But if it wants thousands of independent operators to behave like one recognizable system, it must decide what cannot fragment: the brand promise, operating interfaces, selected standards, sales-linked economics and, in conventional franchises, important rights over the site.
Scale becomes more durable when the organization knows which work can be decentralized and which rights must remain coherent.
This is not a recommendation to franchise. It is a diagnostic question for any network business: what can move outward without causing the system itself to break apart?
AI & Future Lens
AI can make a franchise system more measurable: demand forecasting, labor planning, equipment monitoring, personalized offers and operating assistance can all become more centralized. That creates a paradox. Better central intelligence can improve local execution, but it can also move more decisions toward the center.
The future question is therefore not simply how much AI McDonald’s uses. It is whether AI changes the control boundary between corporation and operator — who decides, who pays, who owns the data, who is accountable and who captures the gain.
Build From This
- Control-right map: list the decisions, assets, data, standards, pricing rights and customer interfaces that sit centrally versus locally.
- Operator-health dashboard: monitor whether partner economics remain strong enough to fund required reinvestment.
- Boundary review: when new technology arrives, explicitly decide whether the new capability belongs at the center, at the edge or in a shared layer.
Remember This
- Franchising decentralizes execution; it does not eliminate control.
- Conventional franchises, developmental licenses and affiliates do not share the same capital structure.
- Real estate is a major economic layer, not the whole company.
- Franchisee health is part of McDonald’s own system health.
- The transferable lesson is to design the boundary between local decisions and central rights.
Primary sources
Facts, figures and quotations should be traceable to the sources below. Sidy's synthesis is labeled as synthesis and does not replace sourced facts.
