Physical restaurant vs dark kitchen: definition and income structure

What exactly is a dark kitchen versus a physical restaurant?

A dark kitchen is a production kitchen with no dining room, built solely for delivery and pickup, while a physical restaurant combines the kitchen with on-site service space, a bar and sensory experience.

That structural difference — not food quality — is what defines both models, and it's where most of the sector's confusion starts. A physical restaurant sells food PLUS ambiance, location, front-of-house staff and street-level brand visibility; a dark kitchen sells only delivered food, with the Glovo, Deliveroo or UberEats algorithm deciding how many people see it. Spain's Hospitality Report 2026 shows 34% of new food businesses in the country now launch as dark kitchens, up from just 8% in 2020. This isn't a passing trend: it's a real shift in how a food business gets validated before committing capital to a location.

Fixed cost structure separates the two models from the first euro

A mid-neighborhood physical restaurant carries rent (8-15% of revenue), utilities (3-5%) and front-of-house labor (12-18%) — a fixed floor of €6,500-8,500 monthly before any margin exists. A dark kitchen, by contrast, pays kitchen rent (3-7%), utilities (1-2%) and only kitchen/packing labor (6-10%), cutting its fixed floor to €3,200-4,800 monthly. That gap explains why validation timelines differ so sharply: physical demands 6-9 months and €45,000-80,000 in investment to know if it works, while dark kitchen answers in 2-3 months with €8,000-15,000. Neither model is inherently better; one risks capital faster and the other protects margin once established. Diego F. Parra, through Masterestaurant's audits, stresses that confusing 'cheap to launch' with 'cheap to sustain' is the first mistake of anyone who has never operated both formats.

Break-even point isn't measured the same way in both models

A physical restaurant's break-even arrives at 65-75% sustained occupancy during peak hours; a dark kitchen's arrives at 2,500-3,200 monthly orders in a major city. These are two distinct calculations because the variables driving them differ completely: occupancy depends on tables, shifts and front staff; volume depends on platform algorithm position and delivery radius. A restaurant averaging 42% occupancy already sits below the critical threshold for positive EBITDA, and a dark kitchen that never touches 2,500 orders/month collapses in 8-12 months, a pattern that repeats across the sector. The trap is comparing both figures as if they measured the same thing — they don't, and treating occupancy as if it were volume, or the reverse, produces cash projections that never hold up.

How this plays out in practice: a case with full numbers?

Take a dark kitchen launched with €12,000 investment and 3 kitchen staff: month 1 closes at 1,100 orders and €3,200 revenue;

month 3 climbs to 2,800 orders and €8,400 — already above minimum viable volume. That accelerated growth rhythm is typical of the virtual model, but so is its fragility: if the platform algorithm trims the delivery radius, the drop can hit 300 orders in a single week. A 65 m² physical restaurant with 20 seats, meanwhile, opens at 45 covers/night in its first month — a lower revenue figure than the comparable dark kitchen — but reaches a stable 90 covers/night eight months later, with margin rising through repeat business. The operational reading is simple: dark kitchen grows fast or disappears; physical grows slow but stays, as long as occupancy holds.

What are the most common errors in interpreting both models?

The most common error treats 'dark kitchen' and 'unbranded virtual restaurant' as synonyms for an inferior business, when in reality it's a model with its own math, not a stripped-down version of physical.

The second, equally widespread error assumes volume validated in dark kitchen translates 1:1 into dine-in covers once a location opens — it doesn't, because someone ordering through an app and someone sitting down to eat respond to different triggers, price and speed versus ambiance and service. The third, costlier error calculates a physical restaurant's break-even using dark kitchen order logic, ignoring that peak-hour occupancy is the variable that actually governs a dining room. Masterestaurant sees this misunderstanding often in operators migrating from one model to the other without adjusting their spreadsheet to the correct cost structure.

Platform commission changes margin defense entirely

In dark kitchen, the Glovo, Deliveroo or UberEats commission — 15% to 30% of the ticket — isn't negotiable; it's the fixed price of existing on that channel, and it cuts net margin by an average 22% versus catalog price before commission. In physical, the equivalent is front-of-house staff, costing 12-18% of revenue, but that person defends brand and retention in a way no platform does for a dark kitchen. The metric that decides which weighs more: if your dark kitchen retention drops below 20% annually, platform commission ends up costing MORE than paying a server who builds loyalty. Annual retention runs around 60% in brand-consolidated physical restaurants, against just 25-35% in unbranded dark kitchens — the difference that truly protects margin over the long run.

When does it make sense to move from dark kitchen to physical restaurant?

The direct answer: when dark kitchen customer retention exceeds 30% by month 2 without a discount, because that figure signals real product traction rather than opening-week curiosity.

Before that point, opening a physical location with €45,000-80,000 in investment repeats the same underlying problem — lack of market — with capital multiplied fivefold or more. If retention instead sits below 20%, the right move isn't jumping to an expensive physical restaurant, but opening a second dark kitchen in another neighborhood, which costs a fraction and spreads the risk. This sequence — validate cheap, scale expensive only with evidence — is the reading that separates an operator with judgment from one confusing enthusiasm with proven demand, and it's the foundation of the hybrid model Masterestaurant recommends from month 12 onward.

The hybrid model resolves the tension between brand and volume

Physical restaurant plus satellite dark kitchen — second brand, second neighborhood, delivery-only — is the most robust model after 12-18 months of operation, because it combines what each format does best: physical delivers brand and 60-75% retention, dark kitchen delivers volume and reach with no space ceiling. The apparent tension — invest in experience or in speed — gets resolved not by picking a side but by sequencing: validate each model separately first, then blend them once both sets of data are in hand. The error that ruins this move is mixing too early, when neither physical retention nor dark kitchen volume has been proven, because then you don't know which of the two businesses is failing or why. Diego F. Parra notes the hybrid isn't a shortcut: it's the payoff for having done the validation work properly in each model on its own.

5 differences that move operating margin

BREAK-EVEN POINT. A mid-neighborhood physical restaurant needs €9,500-11,500 monthly revenue just to cover rent, utilities and front staff — that's the fixed floor, before margin. A dark kitchen with the same revenue volume pays €3,200-4,200 fixed and covers variable costs faster. Physical wins on retention; dark kitchen on break-even speed. PLATFORM COMMISSION vs LABOR COST. In dark kitchen, that 15-30% Glovo/Deliveroo/UberEats commission is non-negotiable; it's the price of reach. In physical, you pay front-of-house (server + host) at 12-18% of revenue, yes, but that person defends brand and retention. The metric: if your dark kitchen retention is below 20%, platform commission costs MORE than paying a server. OCCUPANCY vs VOLUME. Physical breathes with occupancy: 80% occupancy at peak is profitable; 50% is slow bleed.

5 differences that move operating margin — in practice

Dark kitchen doesn't know occupancy; only order volume. That sounds better — no customer, no cost — but platform algorithm volatility compensates: one delivery radius drops (3-4 km) and you lose 300 orders/month overnight. MARGIN DEFENSE. In physical, customer pays for brand and experience — gross margin resists volume drops because price stays up. In dark kitchen, margin is defended ONLY by volume and low price; if volume drops, the only lever is lower price, which kills margin. No experience to protect it. MONEY HORIZON. Physical restaurant: €45,000-80,000 investment, break-even month 10-15, EBITDA positive month 18-24. Dark kitchen: €8,000-15,000 investment, break-even month 3-4, EBITDA positive month 7-9. Physical costs more and takes longer, but more defensible; dark kitchen faster and more fragile.