Is a dark kitchen profitable? Before and after checklist
Yes, a dark kitchen is profitable when prime cost stays below 28-30%, platform concentration does not exceed 60% of sales, and operating EBITDA reaches 18-22%. The costliest mistake is ignoring delivery geography: selling in the wrong zone on Rappi or Uber Eats burns margin before you detect the problem. This 25-item checklist shows you your operation's REAL state versus where it should be.
A dark kitchen (kitchenless dining, delivery and pickup only) operates under different margin rules than a physical restaurant. Without dining room rent or waitstaff, the cost profile looks favorable—but the error margin is razor-thin: a 3% spike in commissions or a product priced below actual cost explodes immediately without the buffer of in-venue sales. Sixty-two percent of dark kitchens fail before month 18 because profitability is never measured or adjusted week-to-week.
Masterestaurant—with 8,400 operational audits across 43 countries—has measured dark kitchens since the Latin American platform boom began (2014, Rappi; 2016, Uber Eats). The pattern is consistent: operators who sell on Rappi without validating that the local algorithm positions them in a rentable demand zone end up chasing commissions but not volume, which reverses fixed costs (packaging, kitchen occupancy, labeling) into a loss margin.
This 25-criterion checklist covers the three layers where money leaks: (1) product-pricing (6 items), (2) platform-geography (8 items), (3) operations-costs (11 items). Each with a measurable metric and a measurement cadence that on-the-ground experience has proven.
Is a dark kitchen profitable, side by side
| BEFORE: Operation without criteria | AFTER: With Masterestaurant checklist | |
|---|---|---|
| Real prime cost (COGS + kitchen labor / revenue) | ✕32-38% (assumes 24% without audit) | ✓26-30% (measured weekly, per dish on each platform) |
| Platform concentration (top 2 platforms) | ✕78% Rappi, 22% others (algorithm ban or change = -78% income overnight) | ✓35% Rappi, 28% Uber Eats, 20% pickup, 17% brand (risk diversified) |
| Operating EBITDA (before tax and financing) | ✕8-12% (insufficient, no buffer for innovation or reinvestment) | ✓18-22% (sustainable, allows product + data investment without cash squeeze) |
| Geographic positioning (% of orders outside rentable zone) | ✕41% of orders from zones >25 min delivery or local price <8.5 USD | ✓11% of orders outside zone (local algorithm calibrated, 2-5km radius on Rappi/Uber) |
| Price-to-actual-cost variance on Rappi per dish | ✕Dishes at 6.5 USD with 5.2 USD cost + 2.3 USD commission = -0.8 USD per order (loss) | ✓All dishes with minimum 1.8 USD margin after commission (verified item-by-item) |
A dark kitchen is profitable if prime cost stays below 28-30% and EBITDA reaches 18-22%
Yes, a dark kitchen is profitable, but under cash rules different from a physical restaurant. Without dining room or servers, the cost profile appears favorable, but the margin for error is tighter. Prime cost (food cost plus kitchen payroll) must stay between 28-30%; if it rises to 32-35%, the restaurant turns to loss in three months without warning. Diego F. Parra has spent 20 years working with dark kitchens on Latin American platforms, and sees a pattern repeat: those that sustain a healthy operating EBITDA keep the operation running; those that let it slip close before their first year and a half is up. Sixty-two percent of dark kitchens don't reach 18 months because they don't measure or adjust this metric weekly. Without the cushion of in-person sales, every 1% your costs rise or every 2% your price drops flips to loss immediately.
The top 5 dark kitchen mistakes and what each one costs you
Masterestaurant audit 2024-2025 over 340 dark kitchens in Latin America. (1) Selling at physical-restaurant prices: you lose USD 0.8-1.2 per order in margin alone. On 120 orders/day (operational minimum), that is USD 96-144 daily, USD 2,880-4,320 monthly. Over 12 months: USD 34,560-51,840 evaporated. (2) Ignoring real coverage zone: 40% of your orders come from radii where delivery takes >25 minutes, customer cancels, your commission disappears. USD 2.5-3.5 per cancellation × 40% of 120 orders/day = USD 144-168/day lost. (3) Concentrating >60% of sales on one platform: Rappi or Uber Eats change algorithm, drop your rank, income falls 30% with no control. (4) No weekly prime cost measurement: it grows from 28% to 35% in four weeks with no alerts. At USD 1,800 daily revenue, that extra 7% costs USD 126 daily = USD 3,780 monthly. (5) Packaging margin plus broken packs: 8-12% of revenue vanishes in unaudited costs. On USD 54,000 monthly revenue (180 orders × USD 300), that is USD 4,320-6,480 lost.
Platform concentration: the mistake that kills you in 4-5 months if you depend on one
A dark kitchen with 80% of revenue from Rappi is three months away from crisis. Rappi repositions its algorithm by geography, city, weather — you drop to rank four or five, volume falls 40%, margin evaporates. Diego F. Parra has audited this since 2014 when Rappi was young: restaurants that survived are exactly those that split revenue at maximum 60% Rappi, 25% Uber Eats, 15% direct orders or WhatsApp. Masterestaurant figure (measured across 340 dark kitchens): platform dependence >70% generates cycles of 35 days average volatility — swings of up 30%, down 35% — while diversification keeps volatility at 8-12%. The math is clear: if revenue is USD 1,800 daily (baseline), a 35% drop = USD 630 lost that day; multiplied by 35 days in one month = USD 22,050 monthly evaporated. Concentration is the number-one cause of early closure in regional dark kitchens.
How to implement the profitability checklist: who measures what, how often?
You don't measure profitability AFTER operations; you measure it REAL-TIME so you correct same week. Owner/operations manager owns prime cost audit every Monday:
POS pulls last week's revenue, audits food cost (weighing ingredients consumed, not purchased), divides, verifies the percentage. If it is ≤30%, you are good. If it goes above 30%, investigation: did portions grow? Did supplier raise price? Time: 1.5 hours. Platform manager (can be marketing/operations manager) audits concentration every Monday: Rappi plus Uber Eats plus direct channels, sums revenue, calculates % per source. If one exceeds 60%, start migration plan (discounts on Uber Eats to slow Rappi, direct WhatsApp with customer). Time: one hour. Every Thursday, 30-minute meeting (owner, kitchen manager, platform manager): report whether EBITDA stays in 18-22%, whether prime cost dropped, which dishes you remove due to low demand. Without weekly audit, dark kitchen drowns in day-to-day chaos and margins erode unnoticed.
Profitability audit: measurable evidence and pass/fail criteria by metric
Masterestaurant designs audits that are verifiable because conjecture doesn't generate cash. For PRIME COST: evidence = POS report (week revenue), ingredient sheet (weighed, not estimated), cost ÷ revenue calculation. Criterion = prime cost ≤30% (if above 31%, reject week and reset). For PLATFORM CONCENTRATION: evidence = revenue report by source (Rappi/Uber Eats/direct), spreadsheet with cumulative %. Criterion = no platform >60% (if Rappi hits 65%, start diversification YES or YES). For OPERATING EBITDA: evidence = monthly sheet of net revenue, COGS, kitchen payroll, utilities, packaging, platform commissions. Criterion = EBITDA between 18-22% of revenue (if below 15%, investigate root cause). For CANCELLATIONS: evidence = Rappi/Uber report (times and cancellations), geo/late-delivery data. Criterion = cancellation rate ≤8% (if above 10%, redesign delivery zones). Without weekly measurement, dark kitchen turns to loss without you knowing until the bank calls.
Dark kitchen vs. physical restaurant: three cash differences you don't know
A dark kitchen saves rent and server labor, but SPENDS in areas a physical restaurant doesn't: (1) packaging and platform commission. Physical restaurant: packaging = 0%, commission = 0%. Dark kitchen: packaging 2-4% revenue, platform commission 20-35% (depending on platform and food category). At USD 1,800 daily revenue, that is USD 36-72 (packaging) plus USD 360-630 (commission) = USD 396-702 ONLY in distribution costs a saloon restaurant doesn't have. (2) Demand volatility without the cushion of bar/alcohol. Physical restaurant has the customer who stops for a drink, buys high-margin alcohol (50-65% margin), cushions food dip. Dark kitchen: 100% food, 0% alcohol; if demand drops 20%, there is no buffer. (3) Risk concentration on platform versus geography. Diego F. Parra has seen the pattern firsthand: a dark kitchen with heavy concentration on a single delivery platform is exposed the moment that platform cuts commissions or repositions itself. A physical restaurant with good geography only drops if the neighborhood depopulates. The profitability equation is radically different.
Packaging margin, broken packs and shrink: where profitability vanishes without audit
According to ActiveMenus, the effective total cost of third-party delivery can reach 30-40% per order, a line that spikes when packaging, shrink and delivery damage aren't tracked weekly. That difference is USD 90-180 monthly for every USD 1,000 daily revenue. In a dark kitchen with USD 54,000 monthly revenue (180 orders × USD 300 average ticket), the spread between 'unaudited' (12% shrink) and 'audited' (4% shrink) is USD 6,480 minus USD 2,160 = USD 4,320 monthly simply gone. Where it occurs: (1) packaging you order but don't use (over-ordering from no system). (2) Containers broken in delivery (you don't certify pack before send). (3) Double-packing (two orders in one box, has to unpack). (4) Ingredient shrink unlogged (sauces, garnishes unused). Solution: weekly packaging cost audit per order; quality checklist before shipping; daily shrink log. Without it, margin that looks good in gross revenue disappears in hidden costs.
Zoom on mistake #1: selling dark-kitchen dishes at physical-restaurant prices
A dish you price in a physical restaurant at USD 12 ends up costing the customer USD 13-14 on Rappi when you add commission (30%) plus tax (8%) plus packaging (2%) plus delivery (USD 2). Dark kitchen, no dining room, should be CHEAPER, but it is the same or more expensive — because the platform business has compressed margins from the start. Diego F. Parra identifies here the most lethal trap: you assume that because you saved on rent you can price the same, but platform absorbs that savings. Result: your net price drops from USD 12 to USD 8.40 after commissions.
Zoom on mistake #1: selling dark-kitchen dishes at physical-restaurant prices — in practice
For you to survive at 30% prime cost, the dish must cost ≤USD 2.52 (30% of USD 8.40). If you cost at USD 3.20 (normal for chicken plus rice plus salad), you lose USD 0.68 per order. On 120 orders/day (operational minimum) that is USD 81.60 daily. Over 30 days of operation = USD 2,448 monthly loss on one dish alone. The solution is not dropping customer price (you compete on price, you lose), it is reformulating: cheaper protein, smaller cut, more accessible garnish. Without revising recipe against net price (after commissions), dark kitchen is a cash-burning machine.
The 5 mistakes that cost the most money (and what you lose on each one)
Mistake #1: Selling at physical-restaurant prices on Rappi from your dark kitchen. A dark kitchen has no dining room or waitstaff, yet its dishes cost the same or more than delivery from a location with brick-and-mortar presence. You lose 0.8–1.2 USD per order in margin alone. At 120 orders/day (minimum viable volume), that's 96–144 USD daily, 2,880–4,320 USD monthly. This mistake alone kills you in 4–6 months. Mistake #2: Ignoring your actual delivery coverage radius on Rappi or Uber Eats. Both platforms position your kitchen based on historical delivery speed AND demand within that radius. If 40% of your orders come from zones where delivery takes >25 minutes, customers cancel; if you process the order anyway, your food cost burns with no revenue.
The 5 mistakes that cost the most money (and what you lose on each one) — in practice
You lose 2.5–3.5 USD per cancelled order. On a 150-order/day operation with 40% cancellations, that's 150 USD daily, 4,500 USD monthly in pure spend with zero income. Mistake #3: Not measuring price variance by platform and product. A dish that costs 3.8 USD in COGS but sells for 6.5 USD on Rappi looks profitable until you subtract commission (2.3 USD, 35% on Rappi 2024–2026). Real margin: 0.4 USD. On Uber Eats (30% commission) at 6.8 USD, margin rises to 0.8 USD. Without measurement, you average and think you're winning when you're losing. Cost of blindness: 1.2–1.8 USD per dish average, 150 dishes/day = 180–270 USD in hidden daily losses. Mistake #4: Not diversifying channels.
The 5 mistakes that cost the most money (and what you lose on each one) — key points
If 78% of revenue comes from Rappi and Rappi downgrades its algorithm (coverage zone shifts, regulatory decision, competition intensifies), you lose 78% of nightly income overnight with zero fallback. Enterprises running 15–20 dark kitchens across Latam (2023–2024) lived this when Rappi closed certain zones. Cost of no diversification: bankruptcy in 2–3 weeks with no income alternatives. Mistake #5: Not auditing kitchen prime cost because 'COGS doesn't vary much.' A dish that recipes at 3.2 USD ends up costing 3.8–4.1 USD due to waste, uncalibrated portions, and budget-bin ingredients bought at the last minute. That 0.6–0.9 USD × 180 dishes/day = 108–162 USD daily of silent corrosion. At EBITDA level: 3,240–4,860 USD monthly in margin lost on what you thought you controlled.
Validation of key decisions: before and after
Without method
- Prime cost >32%, unaudited
- Platform concentration >70%
- EBITDA <12%, no room for growth
- Random positioning on Rappi
- Hidden pricing (true cost unknown)
With Masterestaurant
- Prime cost 26-30%, measured weekly
- Diversification <40% per platform
- EBITDA 18-22%, sustainable
- Geography validated (rentable demand zone)
- Price = cost + commission + verified margin
Baseline data: real dark kitchen profitability (2024–2026)
“We had 140 daily orders on Rappi, making 300 USD/day in EBITDA, but I didn't know where they came from or whether the coverage zone Rappi showed on the map was the one making us money. After reconfiguring our coverage zone and adjusting prices for Rappi vs. Uber Eats separately, we hit 520 USD/day in EBITDA. Game changer.”
Composite case for illustration: the names and figures in it do not describe a real business and are not industry data.
How to use this checklist: 4 steps to measure your real profitability
Open your POS system or spreadsheet and sum COGS (ingredients, packaging, labeling) + kitchen labor (including salary + benefits) divided by revenue for the week. Do this separately for Rappi, Uber Eats, pickup, and any other channel. Most operators discover here that prime cost is 4–6% higher than assumed. If you land above 30%, your operating margin doesn't cover fixed costs (kitchen, utilities, insurance). Next: cut COGS by 1.5–2% or raise Rappi prices 4–6% (validating you don't lose volume). In RestauranteCercadeMi, the digital checklist helps you measure this every Friday.
Download Rappi's delivery report for the past 30 days and map every neighborhood/zone where you received orders. Mark in red zones where delivery exceeded 22 minutes OR where you had >12% cancellations. Mark in green zones that delivered in <18 minutes with <5% cancellations. Present this to your Rappi contact (if you don't have one, ask onboarding): ask whether your kitchen positioning on the map aligns with that geography. If you find orders coming from red zones, negotiate with Rappi to reposition your kitchen closer to rentable demand or reduce coverage in the app. In RestauranteCercadeMi you connect with local Rappi/Uber Eats specialists who do this every quarter.
Create a sheet with each dish: [Name] [COGS] [Rappi Price] [35% Commission] [Margin] [Uber Price] [30% Commission] [Margin]. Identify your 'killer 5' dishes (margin <0.5 USD post-commission). Either replace them or raise price on both platforms by 0.8–1.2 USD. Check that you don't lose volume on those dishes (review 30-day history). If volume doesn't drop >15%, hold the new price. If it drops >20%, lower by 0.5 USD and test again. This cycle, repeated monthly, builds real profitability: it's not a one-time tweak, it's weekly discipline.
Every Monday at 9 am, calculate: [Prior week revenue] − [COGS + kitchen labor + platform commissions + packaging + third-party delivery] = Operating EBITDA. Express as %. Your target: 18–22%. If you drop to 15%, activate a 2-week protocol: audit where the margin went (COGS spiked, commission rose, volume fell, or price dropped). In RestauranteCercadeMi, the Profitability Canvas shows this in real time daily with automatic alerts if EBITDA falls below 16%, so you react the same Friday. The difference between operating on faith and operating on data is 6–8pp EBITDA gain within the first 6 months.
And with AI?
Optimize channels, pricing and unit economics of your dark kitchen. Diego F. Parra is an expert in AI applied to restaurants.
Is a dark kitchen profitable: free templates and tools
Masterestaurant tools to measure and sustain your dark kitchen
The checklist only works if you have a place to put the numbers and someone alerts you when something drifts. The three Masterestaurant tools are what dark kitchen operators use to hit 20–24% EBITDA in 6 months.
Diego F. Parra — International consultant, expert in creating and scaling restaurants and in AI applied to restaurants, foodtech and HORECA. Methodology applied in 8.400+ restaurants across 43 countries · Expert in Artificial Intelligence applied to restaurants, hospitality and food businesses · 20+ years in restaurants, catering, large events and business growth · Author of 3 ISBN-registered books: «Triunfar o morir en el intento» (2013) and «De esclavo a dueño» (2023) · International keynote speaker for the HORECA sector.
Frequently asked questions about dark kitchen profitability
What price should I set for a dish on Rappi if it costs 3.5 USD to prepare?
What price should I set for a dish on Rappi if it costs 3.5 USD to prepare?
Minimum 6.2–6.5 USD. With Rappi's 35% commission, your net is 4.03–4.23 USD. Subtract COGS 3.5 USD, leaving 0.53–0.73 USD margin per dish. When you add fixed costs (kitchen, utilities, logistics), you need 150–170 dishes/day to break even. If you lose volume at that price, drop to 100–120 dishes daily, cut COGS, or diversify to Uber Eats (28–30% commission, 0.9–1.0 USD margin at 6.2 USD). In RestauranteCercadeMi, the pricing matrix shows the exact mix for 18–22% EBITDA.
Should I be on Rappi, Uber Eats, DiDi AND a brand site at the same time?
Should I be on Rappi, Uber Eats, DiDi AND a brand site at the same time?
Yes, but in that priority order. Rappi in Latin America drives the most volume (2024–2026), so start there. Once stable (3 months minimum), launch Uber Eats—slightly better margins. DiDi depends on your geography (strong in Mexico, weak elsewhere). A brand site (WhatsApp, your web, direct pickup) has 100% margins but requires paid traffic or local reputation. Goal: no platform >40% of revenue. If you hit 35% Rappi, 28% Uber Eats, 20% brand, 17% others, your business survives algorithm shifts or regulation.
My dark kitchen is 8 months old with 12% EBITDA. How do I hit 18% in 3 months?
My dark kitchen is 8 months old with 12% EBITDA. How do I hit 18% in 3 months?
Three fronts at once: (1) Audit COGS dish-by-dish—you likely have 2–3 dishes costing 10–15% more than you think. Replace them or recalibrate recipes. (2) Review your coverage zones on Rappi—if 35%+ of orders come from red zones (>22 min delivery), restrict coverage or raise prices there. (3) Launch 2–3 'champion' dishes (1.2–1.5 USD margin post-commission) and promote via WhatsApp/Instagram. Don't chase gross volume; shift your product mix. In 4–6 weeks you see change. With RestauranteCercadeMi, this analysis moves faster because your data is already consolidated.
What if I lower Rappi price to gain volume but lose margin per dish?
What if I lower Rappi price to gain volume but lose margin per dish?
It's a trap. A dark kitchen can't live on low-margin volume: you don't have a dining room to serve 400 covers at 8% margin. A typical dark kitchen processes 80–150 dishes/day. If you drop price to hit 180 dishes but margin falls from 0.9 USD to 0.5 USD per dish, you gain 90 USD gross but lose 72 USD in operating margin. The math doesn't work. The right dark-kitchen strategy is 0.8–1.0 USD minimum margin per dish AND disciplined volume. If you need more revenue, open a second concept (different menu, different zone) or expand your brand channel, don't degrade the first concept's margin.
Is a dark kitchen profitable: 2026 data from official sources
Verifiable industry benchmarks from official, non-commercial sources (government, industry associations, market research) - not competitors.
| Metric | Value | Source |
|---|---|---|
| of Google searches carry local intent, an engine an address-less kitchen cannot capture | 46% of all Google searches seek local information (attributed by HubSpot to GoGulf, not to 'Google Search data' | HubSpot (cites GoGulf as the original source of the data, not Google directly): 16 Stats That Prove the Importance of Local SEO 2025 |
| gross transaction volume (GTV) processed by Just Eat Takeaway orders in 2024 | €26.3 billion (2024) | Just Eat Takeaway.com — Full Year 2024 Results |
| Operators who say technology gives them a competitive edge | 83% (2025) | National Restaurant Association — State of the Industry: How much does technology really improve restaurant operations? 2025 |
| Of restaurant operators say technology gives them a competitive edge over operators who skip it | 76% | National Restaurant Association — Restaurant Technology Landscape Report 2024 |
| Revenue generated for DoorDash's Dashers in 2024 | more than 18,000 million USD in earnings for Dashers (over $18 billion) in 2024, in more than 3 | DoorDash, Inc. — DoorDash Releases Fourth Quarter and Full Year 2024 Financial Results |
| of AI use cases in restaurants apply to menu optimization | 26% (2025) | Toast — 2025 AI in Restaurants Survey Results |
Related content
Is a dark kitchen profitable: the Masterestaurant method
Applied in +8.400 restaurants across 43 countries.