Physical restaurant or dark kitchen: which is best (financial analysis 2026)

Dark kitchen or physical restaurant: which makes sense based on your margin

A dark kitchen is a production space with no dining room that serves only through delivery platforms, while a physical restaurant sells both in-house and online. The difference is not just architectural — it is a margin equation: the physical restaurant tolerates net profitability of 22-28% with proper management; the dark kitchen, selling only on Rappi or Glovo without added structure, lands at 8-12% net margin after commissions that capture 25-35% of each transaction (according to iFood and Delivery Hero 2024 data). The common mistake is thinking «less space equals less cost equals better business»: wrong. It is an equation about where your revenue comes from and how much it leaves you with. The rent savings are real: a physical restaurant on a premium corner pays 6,000+ EUR/month because location is part of your product; a dark kitchen pays 1,800-2,200 EUR/month for logistics warehouse with no visibility, based on Spanish and Italian markets audited by Masterestaurant (2024-2026).

Why rent drops in dark kitchen, but margins do too?

That weekly savings of 1,000 EUR looks decisive until you run the transaction numbers. A physical restaurant serves a 14 EUR dish with 35-42% gross margin;

in dark kitchen that same dish competes in the app against offerings at 10-11 EUR (the perception that «delivery = expensive» hits your price), closing with gross margin near 22-25%. Subtract commission: in the physical space you have 12-14 EUR net margin per dish; in dark kitchen, 2-3 EUR. The rent savings does not close that gap. A common mental trap: believing that because dark kitchen has no dining room or waitstaff, operations are simpler. False. What disappears is customer-service complexity (server schedules, dining room cleaning, table turnover), but another equally severe complexity appears: logistics with third parties. Your food is only 40% of the delivery experience; the remaining 60% is dispatch speed, packaging, courier behavior — variables you do NOT control.

The mistake of confusing operations with customer experience

A customer judges you by the sum. If your burrata comes out perfect but the delivery driver takes 45 minutes en route, your Trustpilot drops. In a physical restaurant, if you fail, you see the angry customer in real time and adjust on the spot; in dark kitchen, you find out later through reviews. That demands a level of quality control that is higher, not lower, to compensate for the variables you lost. A dark kitchen in Madrid with pasta and fried chicken makes 300 weekly orders at 9 EUR average ticket. Gross revenue: 2,700 EUR/week. Food cost: 1,620 EUR (60% COGS, industry standard). Platform commission: 810 EUR (30% of gross revenue). Rent plus utilities: 550 EUR/week (2,200 EUR/month). Result before payroll: −280 EUR weekly. Add one part-time cook (900 EUR/month), and the operation becomes unviable on that expense alone.

Numerical case: dark kitchen with 300 weekly orders

Now the same concept as a physical restaurant: sells 150 delivery orders plus 100 in-house equals 250 transactions per week instead of 300, but the 100 in-house average 16 EUR with no commission, and 150 delivery stay at 9 EUR. Revenue: 2,900 EUR (350 EUR higher). Delivery commission: 405 EUR. Operating margin: superior. Lower transaction count is offset by higher margins in the physical space. The advice I give to new entrepreneurs is direct: do not open a dark kitchen as your first project. A dark kitchen requires that you already master delivery operations, understand platform logistics, have capital to weather low margins, and have operated online before without anxiety. If this is your first business, start with a small physical restaurant — 40-60 m² with a QR menu of 8-12 dishes, not 60 — with dine-in service that leaves you breathable margins while you launch delivery as an extension.

Masterestaurant recommends: if you are new, start physical

Delivery should grow on top of a viable model, not be the viable model. According to Spanish operators' data, dark kitchens opened in 2022-2023 by founders with no prior food operations experience had 68% closure rates within two years; those that were extensions of an already-operating physical restaurant survived in 82% of cases. Experience matters more than concept. Dark kitchen works when you open it as an extension of an already-profitable, recognized physical restaurant. An established chef who launches a sub-brand for a different niche (the physical is fine dining, the dark is fast casual) wins because they reuse shared kitchen space, proven experience, negotiated COGS, and most importantly, verified quality control. Or when a mid-sized group with multiple locations launches delivery-specific brands without cannibalizing the main menu — a pasta house and a quick pizzeria inside the same dark kitchen. That works because you already have staff, operational standards, and financial margin to absorb the launch.

When dark kitchen DOES make sense: extension of established brand?

But if this is your first project, the dark kitchen carries the learning costs of operations, logistics, and quality with none of the steady income a physical dining room would give you.

That is the line: first the profitable model, then optimization. Initial investment to open: 50 m² physical restaurant plus basic equipment equals 45,000-65,000 EUR (build-out, kitchen, tables, licenses). Dark kitchen equals 12,000-18,000 EUR (equipment, minimal permits, conditioning). Dark kitchen looks cheaper. But compare the time to break-even with real margins: the physical reaches operational profitability in 10-14 months if you hit 150+ transactions/week; dark kitchen takes 18-24 months because low margins require very high volume to add up. And if midway you need to pivot — change concept, reposition, add dine-in — a physical space lets you do it; with a dark kitchen in warehouse-only mode without a customer-facing zone, you are trapped.

Investment curve: why physical costs more upfront but closes faster

The flexibility of the physical is money in future scenarios. That is what the numbers don't say at the start but what a restaurant consultant sees repeated: dark kitchen's low upfront investment gets eaten by low margins and lack of exit if the model fails. This argument has limits you need to know. First, in very-high-rent cities (New York, Singapore, Sao Paulo), where a viable physical location costs 12,000+ EUR/month, a dark kitchen may be your only entry door to the business — the analysis changes because physical rent is an unassuming multiple. Second, if your concept is pure delivery (wings, breakfast, desserts) and you already have mature supply chain with vendors, dark kitchen is cleaner. Third, if your target is high-ticket orders (70+ EUR) where gross margins of 50-60% are real, the numbers allow oscillation.

Limits of this analysis: when dark kitchen IS the answer

But for a typical Spanish entrepreneur with traditional or international cuisine concept, 9-15 EUR ticket and 30-35% gross margins at best, dark kitchen is a trap: it looks cheaper, but it kills margin, demands volume that is hard to achieve, and if it fails, there is nothing worth selling as a going concern. **Rent**: physical restaurant pays 30–50% more because location (visibility, foot traffic) is part of the product; dark kitchen pays only logistics warehouse. If the physical is on a premium corner, rent climbs to 6,000+ EUR/month; dark kitchen never exceeds 2,200. **Experience control**: in physical, ALL experience (plates, presentation, service, speed) is yours — if you fail, you see the unhappy customer in real time. In dark kitchen, experience is 40% yours (food) + 60% platform (dispatch speed, packaging condition, delivery driver behavior). The customer judges you by the sum. **Final price**: a dish you sell for 14 EUR in physical, in dark you must sell for 10–11 EUR to compete in the app (the perception of 'delivery = expensive' crushes price).

Operational differences that weigh on margin

With that and 30% commission, you end up at 3–5 EUR gross margin (before variable costs). In physical, the same dish holds price or rises. **Scalability**: physical restaurant caps at 80–120 covers/day (dining room space); if it grows, you open a second location. Dark kitchen scales with volume on platforms — but each platform has a ceiling on visibility in your category (competes with 500+ dark kitchens of fried chicken). Reaching 300 orders/day in dark requires advertising budget within the app (Rappi Ads, iFood Pro) costing 300–500 EUR/month with uncertain ROI. **Resilience**: if the platform changes algorithm, visibility drops (happened with Rappi in 2024). If your physical restaurant loses customers, you see them, talk to them, adjust menu/prices. In dark, you are opaque to the data on why demand fell.