Investment verdict01Is a Ghost Kitchen Worth It? The Verdict Depends on Who Owns the Customer
A delivery-only kitchen can work when food, labor, packaging, and channel fees stay controlled and at least 30%–40% of sales migrate to direct ordering. A model that remains 80%–100% dependent on third-party marketplaces can be busy, well reviewed, and still produce little owner cash.
Demand is real. The National Restaurant Association reported that 37% of U.S. adults ordered restaurant delivery at least weekly in 2025. But delivery demand is not the same thing as attractive restaurant economics. A ghost kitchen removes the dining room, servers, host stand, and expensive frontage; it does not remove food cost, kitchen labor, app commissions, refunds, packaging, utilities, insurance, or customer-acquisition expense.
The strongest versions are not simply “restaurants without seats.” They are compact production systems built around a narrow menu, fast assembly, items that travel well, predictable peak-hour throughput, and a deliberate path from marketplace discovery to owned customer relationships. The weak versions launch too many virtual brands, discount heavily, and treat app sales as if the gross order value were revenue they get to keep.
At $180,000 in monthly sales, reducing the blended order-channel cost from 25% to 15% improves operating cash by about $18,000 per month before any extra marketing expense. That single line can matter more than saving a few thousand dollars on rent.
The hidden asset is not the kitchen. It is the repeat-customer file: email, text opt-ins, direct-order history, loyalty behavior, and neighborhood-level demand data. Build the model around customer ownership, not just lower occupancy cost.
Startup capital02How Much Does It Cost to Open a Ghost Kitchen?
That is a practical U.S. planning range for a shared or turnkey commercial-kitchen launch with equipment, setup, opening inventory, pre-opening payroll, and a real working-capital reserve. A bare-bones test can start near $25,000–$60,000, while a dedicated raw-shell build can push total funding above $200,000 and toward $500,000.
The quoted “cost to open” is often misleading because it mixes two different numbers: the cash needed to pass inspection and accept the first order, and the cash needed to survive the first three to six months. Shared-kitchen rates themselves can look inexpensive; The Food Corridor's survey-based guidance says U.S. shared kitchens commonly charge $15–$45 per hour, with 42% reporting $20–$29 per hour. Yet a high-volume delivery operation may need reserved blocks, dry and cold storage, cleaning time, receiving access, and peak-hour exclusivity. Those extras change the effective occupancy cost.
| Startup item | Low | High | Planning note |
|---|---|---|---|
| Kitchen deposit and first month | $4,000 | $12,000 | Shared membership, turnkey suite, storage, and access deposits |
| Permits and professional fees | $2,500 | $8,000 | Entity, local licenses, health review, food-safety training, plan review |
| Equipment and smallwares | $12,000 | $45,000 | Supplemental cooking, refrigeration, prep, shelving, utensils |
| Ventilation, electrical, plumbing modifications | $0 | $30,000 | Zero in a suitable turnkey suite; material in a second-generation space |
| POS, order aggregation, printers, and IT | $2,000 | $8,000 | Tablets, kitchen display, routers, printers, setup, integrations |
| Brand, menu photography, onboarding | $4,000 | $12,000 | Identity, packaging files, photography, listings, launch creative |
| Opening food and packaging inventory | $5,000 | $14,000 | Ingredients, disposables, labels, tamper-evident supplies |
| Insurance and utility deposits | $2,500 | $7,000 | General liability, property, workers' compensation, deposits |
| Pre-opening payroll and training | $6,000 | $18,000 | Recipe testing, batch training, mock service, manager setup |
| Working-capital reserve | $25,000 | $60,000 | Usually two to four months of fixed cash burn during ramp |
| Total practical launch funding | $63,000 | $214,000 | Planning range, not a national average |
A dedicated build-out can add roughly $140,000–$300,000 for mechanical work, hood and fire suppression, grease management, refrigeration, flooring, walls, and contractor contingency. That takes the modeled total to about $203,000–$514,000. The right number depends less on square footage than on whether the space already has compliant food-service infrastructure.
Fund storage, refrigeration, packaging, and working capital before decorative improvements. Customers never see the kitchen. They do notice missing items, leaking containers, late orders, and inconsistent food.
Facility choice03Which Launch Model Is Cheapest: Shared Kitchen, Turnkey Suite, or Dedicated Build-Out?
The cheapest opening is not automatically the lowest-cost operating model. Hourly shared space minimizes capital but can become expensive when the business needs 300 or more production hours per month, dedicated storage, early-morning prep, and protected dinner capacity. A turnkey private suite costs more each month but gives the operator control over line layout, receiving, cleaning, and throughput. A dedicated lease has the highest capital burden but can be rational once order density is proven.
Shared commissary
$25K–$90KBest for validating one menu, limited service windows, and under roughly 80–100 orders per day. Watch hourly overages and storage fees.
Turnkey private suite
$63K–$214KBest for a serious launch with reserved capacity, multiple dayparts, and 100–220 daily orders. Usually the best risk-adjusted first unit.
Dedicated build-out
$203K–$514KBest only after demand is proven or when several established brands can share production, purchasing, and management.
The economic crossover should be modeled as effective occupancy cost per order, not simply monthly rent. Include base rent or membership, overage hours, storage, common-area charges, waste, cleaning, receiving fees, parking, delivery-driver congestion, and lost sales when the line cannot access equipment at peak time. A $7,000 suite that supports 6,000 monthly orders costs $1.17 per order; a $4,000 shared arrangement that caps the operation at 2,500 orders costs $1.60 per order before overages.
Do not sign a long lease because the concept has good app reviews. Sign it after the model shows repeat demand, stable prep times, and enough contribution dollars per order to absorb fixed occupancy. Reviews prove product-market fit; they do not prove lease affordability.
Opening sequence04How Do You Open One Without Burning Six Months of Cash?
A disciplined launch takes roughly eight to sixteen weeks in a suitable commercial kitchen. The schedule can be shorter for an existing licensed operation adding a virtual brand, and much longer when plan review, construction, hood work, or zoning is involved. The FDA notes that food businesses face a mix of federal, state, and local requirements, and that permits vary by product and facility; the correct starting point is the FDA's food-business regulatory overview and the local health authority.
Prove neighborhood demand — weeks 1–2, $2,000–$6,000Test a focused menu through pop-ups, catering drops, preorder windows, or an existing licensed kitchen. Measure paid conversion, average order value, repeat intent, and delivery radius.
Lock the model and facility — weeks 2–6, $5,000–$20,000Form the entity, obtain insurance quotes, review the kitchen agreement, confirm storage and receiving rights, and map every required permit before paying a nonrefundable deposit.
Install only what changes throughput — weeks 4–8, $15,000–$75,000Prioritize refrigeration, hot holding, prep capacity, labeling, shelving, and reliable order routing. Buy used stainless and prep equipment when service history is clear; be more cautious with refrigeration and mission-critical electronics.
Engineer the delivery menu — weeks 5–9, $3,000–$10,000Cost every recipe, test 20- and 35-minute travel quality, photograph the actual portions, set modifier logic, and remove items that slow the line or generate refunds.
Soft launch and protect runway — weeks 8–16, $25,000–$60,000 reserveOpen limited hours, cap order flow, track promised versus actual prep time, and expand only when the kitchen holds quality during the busiest 30-minute interval.
Licensing usually includes a business license, food-establishment or health permit, manager food-safety certification, employee food-handler requirements where applicable, fire review for cooking equipment, sales-tax registration, and possible sign, waste, grease, or zoning approvals. A home kitchen is rarely suitable for restaurant-style hot food delivery; local rules determine whether any cottage-food exemption applies, and those exemptions usually cover limited low-risk foods rather than full meal service.
Do not order a full equipment package before the health department and fire authority confirm the proposed layout and cooking process. A cheap used appliance becomes expensive when it lacks the required listing, cannot sit under the approved hood, or overloads the electrical service.
Monthly burn05What Does It Cost to Run a Ghost Kitchen Each Month?
The model below shows a mature single-unit operation at $180,000 in monthly gross order sales, a $32 average order, about 188 orders per day, and a 60% marketplace / 40% direct channel mix. It is a planning case, not an industry average. The labor target is deliberately lean but not fantasy: total labor is 24% of sales, below the 31.7% median reported for limited-service restaurants in 2024, while the National Restaurant Association found profitable limited-service operators at a 30.0% median in its 2025 operations analysis.
| Monthly line item | Amount | % of sales | Model driver |
|---|---|---|---|
| Food and ingredients | $50,400 | 28.0% | Recipe cost, yield, waste, purchasing |
| Packaging and disposables | $9,900 | 5.5% | Container mix, bags, labels, cutlery |
| Marketplace commissions | $24,840 | 13.8% | 60% of sales at 23% blended fee |
| Direct payment and ordering fees | $2,160 | 1.2% | 40% of sales at roughly 3% |
| Refunds, discounts, and promotions | $4,500 | 2.5% | Order defects, appeasement, acquisition offers |
| Hourly production labor | $32,400 | 18.0% | Scheduled hours tied to order demand |
| Supervisor and base labor | $10,800 | 6.0% | Manager coverage and minimum staffing |
| Occupancy and storage | $12,000 | 6.7% | Rent, CAM, storage, waste, cleaning access |
| Utilities, maintenance, and repairs | $4,500 | 2.5% | Gas, electric, water, pest, service calls |
| Marketing, software, insurance, G&A | $8,000 | 4.4% | Owned-channel growth, tech stack, accounting |
| Total operating expenses | $159,500 | 88.6% | Operating profit: $20,500 per month |
Where the monthly sales dollar goes
Food and labor dominate, but marketplace fees are the third-largest modeled cost pool.
The first weekly review should be prime cost plus channel cost: food, packaging, labor, marketplace commissions, direct processing, discounts, and refunds. In this case those lines consume 69% of sales before fixed costs. If the operation cannot see those costs by brand, menu item, daypart, and channel, it is managing from the bank balance instead of the business model.
Channel economics06App Commissions, Direct Orders, and the Real Contribution Margin
Marketplace commission is not a generic 30% rule; it depends on platform, plan, order type, and local terms. DoorDash currently publishes U.S. delivery commission tiers of 15%, 25%, and 30%, while Uber Eats announced U.S. marketplace packages effective March 11, 2026 at 20%, 25%–30%, and 30%. These fees may include demand generation, delivery logistics, transaction costs, or membership exposure, but they still come out of the restaurant's order economics.
Base-case order-source mix
The target is not zero marketplace sales. It is using marketplaces for discovery while building a profitable direct reorder path.
Modeled channel cost: 13.8% of total sales for marketplace commissions plus 1.2% for direct processing and order technology.
For a direct order, the restaurant still pays for payment processing, ordering software, marketing, and either customer pickup or delivery fulfillment. Stripe's standard U.S. card pricing is listed at 2.9% plus $0.30 per successful domestic-card transaction, before any ordering platform or delivery-service fee. The point is not that direct is free. The point is that the restaurant can see the customer, control the offer, and often retain materially more contribution dollars.
That is a 23% order-level contribution on a marketplace order at the modeled cost structure. The comparable direct order, using a 3% processing assumption instead of a 23% marketplace fee, contributes about $13.76 before direct marketing and delivery fulfillment. Channel mix is therefore a first-order profit driver, not a marketing footnote.
Price architecture matters too. Delivery prices may need to cover channel costs, but indiscriminate markups can damage conversion and repeat behavior. The better approach is menu engineering: bundles that raise average order value, sides and beverages with strong contribution, fewer low-margin modifiers, and minimum order thresholds for promotions. Measure contribution dollars per order, not only food-cost percentage.
Break-even math07How Many Orders per Day Does a Ghost Kitchen Need to Break Even?
Using $35,300 of monthly fixed costs, a 31% contribution margin, and a $32 average order, the modeled kitchen breaks even near $113,900 in monthly sales, or roughly 3,559 orders per month.
Monthly fixed costs include $10,800 of supervisor/base labor, $12,000 of occupancy, $4,500 of utilities and maintenance, and $8,000 of marketing, software, insurance, and G&A. Variable costs consume 69% of sales, leaving a 31% contribution margin.
At a $32 average order, $113,871 means about 3,559 orders per month, or 119 per day over a 30-day month. Raise average order value to $36 with bundles and the same revenue requires about 105 orders per day. Drop contribution margin from 31% to 26% because app mix or labor deteriorates, and break-even jumps to $135,769 per month, or 141 daily orders at a $32 ticket.
Illustrative 12-month sales ramp
This scenario crosses operating break-even during month 5; cash break-even can come later because startup debt, deposits, and working-capital replenishment still consume cash.
This is why the first three months require a cash reserve even when the concept is promising. The kitchen may be under break-even while payroll, rent, insurance, software, and marketing are fully payable. A founder who funds equipment but not the ramp is effectively betting that launch-week demand will arrive fully formed.
Owner compensation08How Much Can a Ghost Kitchen Owner Make?
Owner income is not gross sales and it is not the operating-profit line. A working owner may receive a manager salary inside labor expense, plus a residual distribution after debt service, taxes, maintenance capital, and working-capital reserves. A passive owner receives no labor-replacement salary and depends entirely on residual cash.
| Scenario | Monthly sales | Operating margin | Annual operating profit | Working-owner salary* | Residual owner cash |
|---|---|---|---|---|---|
| Conservative ramp | $120,000 | 2.0% | $28,800 | $50,000 | $0–$10,000 |
| Base mature unit | $180,000 | 11.4% | $246,000 | $60,000 | $95,000–$125,000 |
| High-throughput unit | $250,000 | 14.0% | $420,000 | $72,000 | $170,000–$220,000 |
*Salary is included in labor expense and assumes the owner replaces a paid general manager. Residual cash is after estimated debt service, maintenance capex, tax reserve, and growth/working-capital reserve; actual tax treatment depends on entity structure and owner circumstances.
In the base case, the model produces $246,000 of annual operating profit. From that, assume roughly $48,000 of debt service, $18,000 of maintenance and replacement capex, and $55,000–$85,000 for taxes, growth, and working-capital protection. That leaves about $95,000–$125,000 of residual cash. Add the $60,000 working-owner salary already counted in labor, and total owner economic benefit is roughly $155,000–$185,000. This is a modeled mature-unit result, not an average-income claim.
Kitchen wages vary sharply by geography and role. For a national reference point, the Bureau of Labor Statistics reported a May 2023 median of $17.20 per hour for restaurant cooks; current local wages, payroll taxes, workers' compensation, overtime, and benefits should replace that figure in the model. A realistic loaded-labor assumption is often 12%–18% above base wages before benefits, but that range is a planning assumption and can be higher in expensive or highly regulated markets.
Pay yourself for the job you perform, then judge the investment on the profit left after replacing that labor. Otherwise a 70-hour workweek can masquerade as a high-return business.
Production moat09Throughput, Prep Time, and Order Density Are the Operating Moat
Delivery economics are won during the busiest half hour, not across the daily average. A kitchen that can theoretically prepare 220 orders per day may still fail if 45 orders arrive between 7:00 and 7:30 p.m. and the line can only finish 28. The result is delayed handoff, driver congestion, cold food, refunds, lower app ranking, and lost repeat demand.
Track by brand, item, daypart, and order size.
Use a lower target for complex, high-ticket menus.
Separate kitchen defects from courier and customer issues.
Researchers studying ghost-kitchen meal delivery emphasize the need to synchronize cooking schedules and dispatch while balancing delivery speed and freshness; the operational problem is described in the Restaurant Meal Delivery Problem with Ghost Kitchens. The practical model should therefore include peak orders per 15 minutes, average prep time, order-ready variance, courier wait, and ready-to-door time—not only total daily orders.
Menu travel score is a financial KPI
A dish that looks profitable at the pass can destroy margin after 25 minutes in a bag. Rate each item on temperature retention, texture, leakage, condensation, portion stability, assembly time, packaging cost, and refund incidence. Remove or re-engineer any item that creates line bottlenecks or post-delivery complaints, even if its theoretical food-cost percentage looks attractive.
Packaging is part of product quality, not just supplies expense. The National Restaurant Association reported that 90% of off-premises customers would likely order a wider variety if upgraded packaging preserved temperature, taste, and quality. That does not justify expensive containers for every item; it supports testing the package against refund rate, repeat rate, and contribution per order.
Add a second virtual brand only when it uses the same prep base, ingredients, and equipment without slowing the first brand's peak-hour line. Revenue that adds a new station, new inventory family, and new training burden may reduce total kitchen profit.
Capital stack10How Should You Fund the Kitchen, and What Will a Lender Ask For?
Match the funding instrument to the asset. Use owner equity for validation, deposits, branding, and the first loss-bearing months. Use equipment financing for identifiable assets with resale value. Use a term loan for durable build-out and equipment. Keep a line of credit or cash reserve for working-capital timing, not for permanently unprofitable orders.
The SBA's 7(a) program can support working capital, equipment, furniture, fixtures, supplies, real estate improvements, and multi-purpose loans, with a current maximum of $5 million under the SBA 7(a) program. The lender still underwrites repayment capacity, owner injection, credit, collateral where applicable, and management experience. A guarantee does not turn a weak unit model into bankable cash flow.
A lender will focus on debt-service coverage after normal owner compensation, not before it. Show the cash flow after replacing the owner's kitchen labor with a market wage. Also show how much liquidity remains after closing. A borrower who contributes every available dollar to equipment may technically open but has no cushion for a slow launch, refrigerator failure, platform payout delay, or food-cost spike.
If the choice is between a larger menu and three more months of runway, buy the runway. The menu can expand after paid demand proves which items deserve space, inventory, and labor.
Control dashboard11Which KPIs Reveal Trouble Before Cash Runs Out?
A ghost kitchen needs a weekly operating dashboard and a monthly financial close. App ratings and gross sales are useful, but they do not tell the owner whether the order produced cash. The dashboard should connect each operating measure to a financial assumption and an action.
| KPI | Formula | Planning benchmark | Decision it drives |
|---|---|---|---|
| Contribution per order | Net order revenue − variable food, packaging, channel, promo, and hourly labor | Target $8–$14; warning below $6 | Pricing, channel mix, menu removal |
| Prime cost | Food cost + total labor ÷ sales | Plan 50%–58%; investigate above 60% | Recipes, scheduling, purchasing |
| Blended channel cost | App commissions + direct fees + delivery subsidies ÷ sales | Target 12%–18%; warning above 20% | Direct-order investment and platform plan |
| Orders per labor hour | Completed orders ÷ production labor hours | 4.5–6.5 for a streamlined menu | Staffing and line design |
| Median prep time | Median minutes from acceptance to ready | 12–18 minutes; track peak separately | Menu complexity and station capacity |
| Refund/remake rate | Refunded or remade orders ÷ completed orders | Target below 2.5% | Quality, packaging, courier handoff |
| Direct-order share | Direct gross sales ÷ total gross sales | 30%–40% by mature stage | Loyalty, CRM, and margin strategy |
| 60-day repeat rate | Customers ordering again within 60 days ÷ first-time customers | Directional target 25%–40% | Product fit and acquisition payback |
| Cash runway | Unrestricted cash ÷ monthly cash burn | Maintain at least 2–3 months during ramp | Hiring, promotions, and capital timing |
The benchmarks above are planning targets, not universal standards. Menu complexity, ticket size, city wages, delivery radius, cuisine, and daypart all change the right range. The important discipline is trend detection. A one-week increase in prep time may be noise; four weeks of lower orders per labor hour, higher refunds, and falling direct share is a financial warning.
Run near 70%–85% during peak. Below that range, capacity is underused; above roughly 90%, small disruptions create queues and quality failures. “Maximum sustainable” means the rate the team can hold without increasing prep-time variance or defects—not a one-time sprint.
Risk and return12What Payback Period Is Realistic, and When Is the Model Not Worth It?
Simple payback divides the initial investment by annual cash flow available to repay that investment. It is useful, but it ignores the launch ramp, seasonality, taxes, financing structure, and replacement needs. For a modeled $140,000 initial investment, steady-state simple payback ranges from about one year to more than five years depending on contribution margin and order density.
| Payback case | Annual cash for payback | Simple payback | Calendar reality after ramp |
|---|---|---|---|
| Conservative | $25,000 | 5.6 years | Often 6+ years; vulnerable to lease renewal and equipment replacement |
| Base | $75,000 | 1.9 years | Roughly 2.5–3.5 years after a six- to twelve-month ramp |
| Upside | $130,000 | 1.1 years | Roughly 1.5–2.0 years if peak capacity and repeat demand hold |
Use cash after maintenance capex and debt service, not EBITDA. Also exclude any owner salary earned for kitchen work; that is compensation for labor, not return on invested capital.
The model is not attractive when the downside case needs perpetual discounts, more than 75%–80% marketplace dependence, a prime cost above 60%, or a lease that requires mature-unit sales before the first month. It is also weak when the menu cannot travel, the average ticket is too low to absorb packaging and delivery economics, or peak demand requires a second line before the first line reaches stable utilization.
Macro demand should not be assumed. As of May 2026, nominal U.S. eating-and-drinking-place sales were up 2.7% year over year, but inflation-adjusted sales were down 0.9%, according to the National Restaurant Association's restaurant sales indicator. That is a reminder to validate local order volume and price sensitivity instead of relying on market-growth headlines.
| Risk | Trigger | Likely financial impact | Control |
|---|---|---|---|
| Marketplace dependency | Direct share stays below 20% | 5–12 margin points lost versus target mix | CRM capture, reorder offers, pickup, catering |
| Peak-hour congestion | Prep time exceeds 20 minutes | Refunds, lower ranking, labor overtime, churn | Capacity caps, menu simplification, station redesign |
| Food and packaging inflation | Combined cost rises 3 points | About $5,400 monthly at $180K sales | Recipe recosting, vendor bids, bundle pricing |
| Brand sprawl | New concepts add unique SKUs and stations | More waste, training, stockouts, and slower tickets | Shared ingredient architecture and brand P&Ls |
| Working-capital squeeze | Sales grow faster than cash reserve | Payroll or supplier stress despite accounting profit | Weekly cash forecast and minimum reserve policy |
