Restaurant Feasibility Study: Demand, Unit Economics and Financing

A restaurant feasibility study tells a lender whether one concept, on one site, run by one operator, can sell enough meals at a supportable check to pay for food, labor, occupancy and franchise fees, and still cover its debt at the ratio the capital source requires. It answers with a defined trade area, a sales forecast by daypart and channel, and a cost model built from quotes and payroll plans, then tests the result against the program's rules.

Three facts frame almost every restaurant loan file:

• Spending is large and still growing. U.S. food-away-from-home spending reached $1,414,967 million in 2025, 33.9% more than in 2019, and took 56.3% of all food spending [1].
• Labor is the largest cost Census publishes. In 2024, payroll, fringe benefits and contract labor equaled 36.7% of revenue at food services and drinking places [15][16].
• Young restaurants close more often. In 2023, restaurants aged one to five years exited at a 13.4% rate, against 6.5% for those 11 years and older [14].

This page sets out what the study must prove, the federal and state data it draws on, and how each financing program reads it. For how FSC approaches the work, see restaurant feasibility study consultant; for a worked example, see the quick-service restaurant portfolio model study.

Every study ends in one of three conclusions: feasible as proposed; feasible with stated conditions, such as a smaller loan, more equity or a different site; or not feasible as proposed, with the changes that would make it work.

What is a restaurant feasibility study?

It is an independent analysis, written for the lender's file, of whether a proposed or existing restaurant can support its financing. USDA's rule describes a feasibility study as an independent consultant's report "evaluating the economic, market, technical, financial, and management feasibility of the proposed project or operation" [24]. Those five parts are a sound outline for any lender.

Federal statistics split the business by service model. A full-service restaurant (NAICS 722511) serves patrons who order while seated and "pay after eating"; a limited-service restaurant (NAICS 722513) serves patrons who order or select items and "pay before eating" [18]. Fast casual restaurants sit in the limited-service code [18]. Shops serving coffee, juices, ice cream or other specialty snacks fall under snack and nonalcoholic beverage bars (722515) [18], and bars under drinking places (722410). The code matters: it sets the SBA size standard, the Census peer group and, in an acquisition, the transaction category.

A feasibility study is not an appraisal. An appraisal values property at a date; a feasibility study tests whether the operating plan repays the loan. SBA's rule says SBA "may require" a feasibility study [20], and many lenders ask for one when a decision rests on projections.

Nor is it a business plan. A business plan is the operator's case for the concept, written to persuade; a feasibility study tests that case for a reader with no stake in the outcome, and it can conclude that the project does not work as proposed. USDA's rule names both, side by side, among the eligible costs of a Business and Industry project [24].

What does a lender need the study to prove?

A lender needs proof that the restaurant's own cash flow will repay the loan under the program's rules. SBA's regulation states the standard: "Loans must be so sound as to reasonably assure repayment" [20].

Exhibit 1. What the lender asks and what the study shows

Source: 13 CFR Part 120 and 121.201 (revised as of January 1, 2026); SBA SOP 50 10 8.1 and Appendix 15; 7 CFR Part 5001; 40 CFR 403.5; 2010 ADA Standards, as cited in each row.

SBA's SOP lists conditions that may lead SBA to request a study. Three fit restaurants closely: market saturation, a "unique market concept", and "significant rapid growth" of the applicant or its affiliate group with rising unseasoned debt [26]. A multi-unit franchisee opening several units at once can meet the last one. For what separates a study a lender can rely on from one it cannot, see what makes a feasibility study bankable.

How large is restaurant demand, and is it still growing?

It is large and still growing in dollars. USDA's Economic Research Service puts 2025 food-away-from-home spending at $1,414,967 million with taxes and tips, up from $1,057,024 million in 2019 [1]. Its share of total food spending was 57.0% in 2019, fell to 51.8% in 2020 and recovered to 56.3% in 2025 [1].

Exhibit 2. Food-away-from-home spending by outlet, 2019 and 2025 ($ millions, with taxes and tips)

Source: USDA Economic Research Service, Food Expenditure Series, nominal expenditures with taxes and tips [1]. Changes and shares calculated by FSC; other outlets not shown.

Restaurants took 71.0% of the total. Limited-service led the growth since 2019, but full-service rose 7.9% from 2024 to 2025 against 3.3% for limited-service [1]. Alcohol sold at eating and drinking places, counted separately, was $129,764 million in 2025, up 32.1% from 2019 [1], which matters for full-service and bar concepts.

Other federal measures point the same way. Personal consumption of food services reached $1,294.2 billion in 2025, 53.2% more than in 2019 [4]. Sales at food services and drinking places were $1,032,727 million in 2025, up 4.9% from 2024 [5], and $807,271 million from January to September 2026, up 4.4% from the same months of 2025 [6].

These are nominal dollars, so they mix price and volume. A pro forma that raises the check with inflation and adds traffic growth on top must show where the extra guests come from. The downside case should also remember 2020, when sales at food services and drinking places fell 16.9% [5].

How does the study define the trade area and daypart demand?

By drive time and by meal. Lunch demand comes from workers and shoppers near the site; dinner and weekend demand come from households within a short drive; a highway site draws from traffic. The study maps each, then counts the households, income and competing seats inside it.

The competitive survey records each competing restaurant's segment, seats, price point, hours and channels, plus announced openings, because saturation is judged inside the trade area, not from state counts.

The study estimates spending potential from the trade area's households and their income distribution, then tests the share the subject must capture.

Exhibit 3. Restaurant density and spending in the ten largest restaurant states

Source: U.S. Census Bureau, County Business Patterns 2023 [9][11] and Vintage 2025 population estimates (July 1, 2023) [17]; USDA Economic Research Service, state food sales per capita [3] and national total [1]. Residents per restaurant, U.S. per capita and payroll per employee calculated by FSC.

These ten states held 55.2% of U.S. restaurant establishments [11][9]. Among the ten, New York had the most restaurants per resident and the highest payroll per employee; Florida and Pennsylvania had the fewest restaurants per resident. The highest per capita sales are in places with heavy visitor and commuter traffic: the District of Columbia ($9,287.37), Nevada ($6,008.24) and Hawaii ($5,571.96) [3]. A study in a tourist market separates resident from visitor demand. State pages, such as restaurant feasibility study in Texas, work through local data.

The federal series used here do not report restaurant sales by daypart or channel. The study builds them from the operator's point-of-sale data for an existing restaurant, from hourly traffic counts and observed competitor volumes for a new one, and from the franchisor's disclosed data where Item 19 provides it. It then sets the mix of dine-in, drive-through, pickup and delivery, because each carries a different check, labor need and commission cost.

The calendar matters too. ERS monthly data show both segments peaking in May 2025 and bottoming in February.

Exhibit 4. Monthly food sales at restaurants, 2025 ($ millions)

Source: USDA Economic Research Service, monthly sales of food with taxes and tips by outlet type [2]. Index calculated by FSC; months sum to the annual totals in Exhibit 2.

May ran 22.0% above February at full-service restaurants and 22.7% above at limited-service restaurants [2]. Months differ in length and in weekends, so the study models monthly sales and checks the weakest months against the payment.

How does the study forecast average check, covers and seat turns?

From the bottom up: covers or orders in each daypart, times an average check, times the number of services. Each input is tested against evidence outside the operator's plan: competitor menus and traffic for check and covers, seats and hours for turns, and federal price series for growth.

Exhibit 5. Sales build for a 150-seat full-service model (model inputs, not market data)

Source: FSC model, test case B in Exhibit 13. Every input is an illustrative assumption for a 150-seat restaurant.

Here the seats turn 10.6 times a week and 1.2 times on a Friday or Saturday night, the dine-in check averages $29.47, and off-premises orders bring 6.3% of sales. Each number is a claim to support. Turns are capped by hours, kitchen output and meal length, and the check comes from a costed menu mix, not a single guess. Check growth is tested against published menu price change, so a forecast that grows faster, or adds traffic on top, must say why. Delivery gets its own line because platform commissions do not apply to dine-in sales; the study takes the rate from the operator's contracts.

How do QSR, fast casual, full-service and bars compare?

They differ in volume, labor intensity and ownership. Full-service restaurants earn more per unit and spend more on payroll; limited-service restaurants are more often run by multi-unit operators; bars are smaller.

Exhibit 6. Restaurant segments in the 2022 Economic Census

Source: U.S. Census Bureau, 2022 Economic Census, EC2272BASIC [7]; 13 CFR 121.201 (revised as of January 1, 2026) [21]. Ratios calculated by FSC. Size standards are average annual receipts.

Full-service and limited-service restaurants earned 46.6% and 44.9% of 2022 receipts in NAICS 7225 [7]. Limited-service averaged 1.63 establishments per firm, the mark of chains and multi-unit franchisees, against 1.03 for drinking places [7]. Self-service technology splits the segments too: 71.8% of limited-service establishments used electronic devices for self-service orders or payment, and they earned 82.8% of the segment's receipts; 48.0% of full-service establishments did [8].

Exhibit 7. Change in establishments, employment and pay, 2019 to 2023

Source: U.S. Census Bureau, County Business Patterns 2019 and 2023, U.S. files [9][10]. Changes and payroll per mid-March employee calculated by FSC.

Payroll per employee rose 26.7% to 31.5% in four years across the segments, while full-service employment fell. A study that carries a pre-2020 labor ratio into a new budget will understate cost. Many units are small: 26.3% of full-service and 33.3% of snack bar establishments had fewer than five employees in 2023 [9]. Another 166,785 nonemployer businesses in NAICS 7225 averaged $55,973 in receipts in 2024 [12].

Exhibit 8. Restaurant establishments by number of employees, 2023 (share of establishments)

Source: U.S. Census Bureau, County Business Patterns 2023, U.S. file, employment size classes [9]. Shares calculated by FSC from mid-March employment.

Limited-service units cluster at 10 to 49 employees (53.0% of establishments); most drinking places run with fewer than ten (66.1%). A staffing plan should land in its segment's band or explain why not.

Independent concepts and restaurants inside larger projects

An independent restaurant has no franchisor data, so its study rests on the operator's record, the costed menu and the competitive survey. A restaurant inside a hotel, an entertainment venue or a mixed-use building draws part of its demand from the host property. ERS puts 2025 food sales at hotels and motels at $38,596 million, up 4.4% from 2019, and at recreational places at $34,370 million, up 9.9% [1], far slower than restaurants. The study separates captive demand from outside demand and tests the outlet if the host underperforms; see the hotel feasibility study and mixed-use feasibility study.

How do restaurant economics differ by state and between metro and rural markets?

Widely, so a national average is a poor stand-in for any one market. The 2022 Economic Census publishes receipts and payroll by segment for every state. Exhibit 9 covers the 19 states with an FSC restaurant state page.

Exhibit 9. Segment economics in 19 states, 2022

Source: U.S. Census Bureau, 2022 Economic Census, EC2272BASIC, state and U.S. records [7]. Ratios calculated by FSC.

Full-service receipts per establishment ran from $968,238 in Wisconsin to $2,215,317 in Nevada, and limited-service from $970,913 in Pennsylvania to $1,594,596 in Arizona. Full-service payroll shares ran from 30.2% (New Jersey) to 38.6% (Colorado), so a budget that borrows one state's labor ratio for another can miss by eight points of sales. State pages, such as Pennsylvania and California, read these figures with local license and demand data.

Rural markets run smaller. Census also publishes each state's figures for counties outside any metropolitan or micropolitan area, where it does not withhold them. Those counties are not the same as USDA's rural areas, which exclude cities and towns of more than 50,000 inhabitants and their adjacent urbanized areas [24], but they show the direction.

Exhibit 10. Receipts per establishment, statewide and outside metro and micro areas, 2022

Source: U.S. Census Bureau, 2022 Economic Census, EC2272BASIC, state records for areas not in a metropolitan or micropolitan statistical area [7]. "Withheld" marks cells Census did not publish. Ratios calculated by FSC.

Outside metro and micro areas, full-service receipts per establishment were 39.7% (New York) to 58.2% (Michigan) of the statewide figure, and limited-service 70.1% (Texas) to 93.0% (Pennsylvania). A rural full-service plan that borrows a metro unit volume can overstate sales by half, the first gap a USDA Business and Industry study for a rural restaurant must close.

How often do restaurants open and close?

Often. From 2015 to 2023, entry rates for restaurant establishments ran from 9.8% to 11.5% a year and exit rates from 8.1% to 10.3%, according to Census Business Dynamics Statistics [13].

Exhibit 11. Restaurant establishment entry and exit, NAICS 7225

Source: U.S. Census Bureau, Business Dynamics Statistics 2023, establishments by 4-digit NAICS industry [13]. Rates as published, rounded.

There were 49,355 restaurant establishment exits in 2023 [13]. Age matters more than the year: establishments aged one to five exited at 13.4%, those aged six to ten at 8.8% and those 11 years and older at 6.5% [14]. A start-up's downside case should therefore carry a slower ramp and a weaker first year than an established unit's, and the SOP tells lenders to review "the number of failed franchisees" for franchise applications [26].

What do restaurant unit economics and prime cost look like?

Prime cost, food and beverage cost plus labor, is the core of every restaurant budget, and the Census expense data publish only half of it. The Census Annual Integrated Business Survey reports labor, occupancy and overhead for the industry, but its expense file does not publish a food and beverage purchases line for NAICS 722. Food cost therefore enters the model as a labeled assumption until menu costing, supplier invoices or the operator's history support it.

Exhibit 12. Expenses as a share of revenue, food services and drinking places, 2024

Source: U.S. Census Bureau, 2024 Annual Integrated Business Survey, AIES00EXP02 and AIES00BASIC, NAICS 722, revenue $984,202,612 thousand [15][16]. Shares calculated by FSC.

Payroll alone was 37.1% of 2024 revenue at full-service restaurants, 27.4% at limited-service restaurants, 28.4% at snack bars and 32.1% at drinking places [15]. Federal law sets a floor under that line: a $7.25 minimum wage [36], and for tipped employees, those who customarily receive more than $30 a month in tips, a cash wage of at least $2.13 with tips credited for the difference [35]. Many states and cities set higher wages, and the study uses the local rate. Chains with 20 or more locations under the same name must also post calorie counts [34].

The study builds each line from the bottom up: a menu mix and costed recipes for food and beverage, a staffing schedule by daypart for labor, the lease or debt for occupancy, the FDD's Item 6 for royalties and marketing fund, and quotes for utilities, insurance, repairs and card fees. Census ratios are a check on that budget, not its source.

Exhibit 13. Two illustrative unit models (model results, not market data)

Source: FSC calculator, test cases A and B (calculator below). Inputs are illustrative assumptions; the rate assumes the 7.00% prime rate in the Federal Reserve's H.15 release of November 10, 2026 [30] plus 2.75 points, inside SBA's 3.0-point cap for loans over $350,000 [20].

The quick-service case passes, but sales could fall only 2.4% before coverage reached 1.15. The full-service case would pass the 1.15 test for a Business Expansion and fails the 1.25 test that Appendix 15 sets for an Initial Acquisition [26]; at that floor, the same cash flow supports a loan of about $1,881,162 on the same terms. The SOP measures coverage as EBITDA divided by debt service [26], which is how the calculator works.

The calculator turns sales, prime cost, occupancy, franchise fees and other operating costs into EBITDA, divides it by the debt service on up to two loans, and tests the result against the coverage floor of the SBA program you choose; it also reports the sales needed at that floor. Its default inputs are illustrative assumptions, not market data, so replace them with the subject's own figures.

What sales does a restaurant need, and which assumptions decide the outcome?

It needs enough sales to cover its fixed costs and its debt at the program's floor, which is well above the point where EBITDA turns positive. Exhibit 14 sets out three thresholds for the two test cases.

Exhibit 14. Break-even points for the two unit models (model results)

Source: FSC calculator, test cases A and B; floors from SBA SOP 50 10 8.1 [26]. Inputs are illustrative assumptions.

With so much cost fixed, small sales changes move coverage a lot: in the quick-service model, sales 7.7% below the model would only just pay the debt.

Exhibit 15. Coverage when one assumption changes (model results)

Source: FSC calculator, test cases A and B, changing one input at a time; the rate case adds 1.00 point to the 9.75% loan rate. Inputs are illustrative assumptions.

Sales and prime cost decide the result. A 5% sales miss, or a 2-point rise in food or hourly labor cost, pushes the quick-service case below 1.15; a 1-point rise in the prime rate does not, because most of a 10-year payment is principal. The study spends its effort where the model is most sensitive, and reports the downside cases beside the base case.

How long does a new restaurant take to ramp up?

No federal series reports how quickly new restaurants reach stable sales, so the ramp is an assumption the study must support and stress. SBA sets the test: a start-up, a business in operation for 1 year or less, must show coverage of 1.15 within 2 years of funding [26]. In the 504 program, if projections show repayment in Year 2 but not in Year 1, the borrower must show liquidity to cover the Year 1 shortfall [26]. The FTC rule asks franchisors to list "additional funds" for an initial period in Item 7, and says a reasonable period is at least three months [19]; that is a floor for the reserve, not a forecast.

Exhibit 16. Ramp-up of the quick-service model under three paths (model)

Source: FSC model. Quarterly sales are shares of the $1,800,000 stabilized run rate in test case A, with the same costs and $172,616.72 of annual debt service. The three paths are illustrative assumptions.

The three paths end Year 2 at similar sales but make three different files. The steady build passes the 2-year test and needs a reserve of $27,012 for Year 1 debt service, on top of pre-opening costs. The opening surge shows 1.50 coverage in its first quarter, yet its Year 2 coverage is lower than Year 1, which is why a study never annualizes a new restaurant's first months. The slow ramp fails, so the loan needs more equity or a smaller amount. The study supports its chosen path with Item 19 data on newer outlets where disclosed and the operator's past openings.

Should the concept be a franchise or an independent restaurant?

It depends on what the operator lacks. A franchise supplies a brand, a tested system and disclosed outlet data, at the cost of fees, supply rules and less control; an independent keeps its margin and freedom but must prove its concept from scratch.

Under the FTC rule, a relationship is a franchise "whatever it may be called" when three elements are present: the business uses the franchisor's trademark, the franchisor exerts or has authority to exert "a significant degree of control" or gives significant assistance, and the franchisee makes a required payment [19]. SBA uses the same definition for its Franchise Directory [26], and a license or co-branding deal can meet it.

Exhibit 17. Franchise or independent: what changes in the study

Source: 16 CFR Part 436 (revised as of January 1, 2026) [19]; SBA SOP 50 10 8.1 [26]; SBA Franchise Directory [27].

The FTC rule has exemptions, and two often reach restaurants. A sale is exempt when the franchisee's initial investment, excluding unimproved land and franchisor financing, totals at least $1,469,600 and the buyer signs an acknowledgment, or when required payments through the first six months of operation total less than $735 [19]. So is a franchisee in business at least five years with a net worth of at least $7,348,000 [19]; the FTC adjusts these thresholds every fourth year [19]. An exempt sale may come without a disclosure document, but if the brand meets the definition, SBA financing still requires a Directory listing [26].

Large firms, a group that includes chains and large franchisees, hold much of restaurant sales. Exhibit 18 groups 2022 restaurant firms by their total receipts.

Exhibit 18. Restaurant firms by size of enterprise receipts, 2022

Source: U.S. Census Bureau, Statistics of U.S. Businesses 2022, U.S. 6-digit NAICS by enterprise receipts size [37]. Shares calculated by FSC.

In 2022, 467 limited-service firms with receipts of $100 million or more ran 18.3% of limited-service establishments, and firms of $10 million or more earned 56.4% of the segment's receipts [37]. Full-service is far less concentrated: firms under $2.5 million earned 40.9% of its receipts [37]. A franchisee competes with these operators for sites and labor.

What should a lender read in a franchise disclosure document?

The lender reads the fee items, the initial investment, the outlet history, the franchisor's financial statements and, where the franchisor makes one, the Item 19 financial performance representation. The FTC Franchise Rule requires the franchisor to furnish the disclosure document at least 14 calendar-days before the buyer signs or pays [19].

Exhibit 19. Franchise disclosure items a feasibility study uses

Source: 16 CFR Part 436 (revised as of January 1, 2026) [19].

Item 19 needs care. A franchisor that makes a financial performance representation must have "a reasonable basis and written substantiation" for it [19]. When it discloses a subset of outlets, it must state how many and what percent "actually attained or surpassed the stated results" [19]. The study compares the subject's site, format and trade area with the subset behind the figure before using it. A franchisor may also give a buyer the actual operating results of the specific outlet being sold [19], which is the starting point in a resale.

SBA adds its own franchise rules. If the applicant's brand meets the FTC definition of a franchise, "it must be on the Directory in order to obtain SBA financing" [26]. SBA's Franchise Directory page shows an edition effective October 28, 2026, alongside Form 2462, the Addendum to Franchise Agreement [27]. Lenders must review credit information such as "the number of failed franchisees" and franchisor projections [26]. A franchisee run by a management company that is, or is affiliated with, the franchisor is not eligible [26]. A franchise development agreement, which earns royalties from other franchisees, is passive and ineligible; area development rights to open units the franchisee owns may be eligible if the applicant and its affiliates are small [26].

Affiliation now turns on ownership. SBA's 2023 rule "removed the principle of control" as a basis for affiliation [26]. Under the SOP, an individual who owns more than 50% of the applicant and more than 50% of another business in the same 3-digit NAICS subsector makes the two affiliates [26]. Every restaurant and bar is in subsector 722, so a multi-unit owner's other restaurants count toward the size standard and toward SBA's $3,750,000 guaranty limit for one borrower and its affiliates [20]. Franchise restraints themselves "generally will not be considered" in deciding whether franchisor and franchisee are affiliated [22]. An applicant that misses its industry standard may still qualify under the alternative size standard of $20 million tangible net worth and $6.5 million average net income [23].

How does SBA treat a restaurant acquisition?

Under SOP 50 10 8.1 Appendix 15, a change of ownership is an Initial Acquisition by default, and that category requires historical coverage of at least 1.25 and equity of at least 10% that cannot be reduced [26]. The lender may not rely on projections to meet the coverage test [26].

Exhibit 20. SBA 7(a) change-of-ownership rules that shape a restaurant purchase

Source: SBA SOP 50 10 8.1, Appendix 15, effective October 1, 2026 [26].

The four-digit test bites in restaurants. A full-service operator (7225) that buys a bar (7224) is not in the same industry group, so the purchase is an Initial Acquisition at 1.25 even though the buyer runs restaurants. The study rebuilds the seller's earnings from tax returns, point-of-sale reports, bank deposits and payroll records, tests add-backs, and models the transition: new management, any franchisor remodel required at transfer, and the lease assignment. The quick-service restaurant portfolio model study works through a four-unit franchise purchase under these rules.

Which acquisition category applies?

Appendix 15 sorts every change of ownership into four categories, and the category sets the coverage floor, the equity and whether the seller may stay [26].

Exhibit 21. Appendix 15 transaction categories in a restaurant purchase

Source: SBA SOP 50 10 8.1, Appendix 15 [26].

Two limits catch restaurant partnerships. A buyer who has not worked in the business for at least 24 months may take less than 50% of the equity in an Owner Buyout; a larger stake moves the deal to Initial Acquisition standards [26]. A selling owner who keeps less than 20% must give a full guaranty for at least two years after final disbursement [26].

Due diligence is set by the Business Purchase Price, which excludes owner-occupied real estate [26]. Above $350,000, the business valuation must come from a "Qualified Source" and be prepared for the lender, not the buyer or seller [26]. At $3 million or more, a Quality of Earnings report adds a cash proof that reconciles bank statements to the income statement and tax return for the trailing 12 months and the last two fiscal years [26]. If the price exceeds the valuation, equity makes up the difference [26]. The lender analyzes three year-end statements plus the current interim period [26]. In a restaurant, where cash sales, comps and owner meals blur the books, the cash proof is where the asking price meets the deposits.

Special purpose property rarely helps. Appendix 15 allows reliance on projections, and a term up to 25 years when 85% or more of project costs are real estate, only for an owner-occupied special purpose property integral to the business; its examples are storage units and hotels [26].

Kitchen equipment also counts for little as collateral, which matters because most restaurant assets are equipment and fixtures.

Exhibit 22. How SBA counts collateral in a 7(a) change of ownership

Source: SBA SOP 50 10 8.1, Appendix 15 [26].

A leased restaurant is therefore rarely fully secured, and the shortfall rule then reaches the owners' personal real estate [26].

How do alcohol licenses change a restaurant's feasibility?

They shape the check, the budget and sometimes the opening date. Alcohol sold at eating and drinking places was $129,764 million in 2025 [1], and the right to sell it is usually a state license layered on a local approval, with a food test, a seat minimum or a population quota attached, as Exhibit 24 shows.

Exhibit 23. Alcohol spending away from home, 2019 to 2025 ($ millions, with taxes and tips)

Source: USDA Economic Research Service, Food Expenditure Series, nominal food and alcohol expenditures [1]. Changes and shares calculated by FSC.

Alcohol at eating and drinking places fell 27.6% in 2020, and the away-from-home share did not pass its 2019 level until 2025, so a concept that leans on drinks needs a deeper downside case. In a change of ownership, a transferable liquor license conveyed in the sale must be secured as collateral [26]; whether it transfers is a matter of state law.

Exhibit 24. Restaurant alcohol licensing in 18 states

Source: state statutes, rules and agency pages as cited in each row; all read for FSC's restaurant state pages.

Three patterns run through the table. Quota states, such as California, Indiana, Massachusetts, New Jersey and Pennsylvania, cap licenses, so a new restaurant may have to buy one or wait for an auction; the price belongs in the budget and the wait in the timeline. Food-share states, such as Arizona, Kentucky, Michigan, New Mexico and Texas, tie the license to the menu, so the modeled sales mix must pass every year. Local-control states put a city or county decision in the path. In Pennsylvania, an economic development license application asks for projected revenue and expenses [67], a small feasibility study for the license itself.

What site and technical issues decide feasibility?

Zoning, the drive-through, the grease interceptor, parking and the lease decide whether a concept fits a site at all. Most of these rules are local, so the study reads the municipal code for the subject parcel; a few federal thresholds apply everywhere.

Exhibit 25. Site and technical checks for a restaurant

Source: 40 CFR 403.5 (revised as of July 1, 2025); U.S. EPA fact sheet EPA-833-F-12-003; 2010 ADA Standards, sections 208.2 and 226.1; 21 CFR 101.11; SBA SOP 50 10 8.1, as cited in each row.

Grease is the restaurant's sewer issue. EPA's 2012 fact sheet cites its 2004 report to Congress: grease from restaurants, homes and industry is "the most common cause (47%) of reported blockages", and estimates grease trap waste and uncollected grease at 800 to 17,000 pounds a year per restaurant [32]. Interceptors "must be designed and sized appropriately" to let grease cool and separate, and must be serviced regularly [32]. Sewer authorities may adopt grease ordinances, set design requirements for grease traps and inspect them [32], so the study prices the interceptor, its pumping and any local permit.

The drive-through is often the reason for the site. The study compares the stacking the code requires with the queue the forecast implies at the peak hour, and checks whether a queue would back into a public street or block parking. A local traffic study may be required. Codes measure the queue in different units: some count cars, some count feet, and some count from the order point rather than the window.

Exhibit 26. Drive-through rules in four local codes

Source: City of Portland Title 33; City of Raleigh and City of Durham unified development ordinances; City of Saint Paul planning memo of March 22, 2024, as cited in each row.

At Durham's 20-foot space length, Raleigh's eight spaces equal 160 feet, close to Portland's 150. Saint Paul's planners found stacking needs "are typically greater than required" by their code [41]. The larger of the code minimum and the forecast queue governs the site plan.

Health plan review comes before construction. In El Paso County, Colorado, new construction or an extensive remodel needs plan review first, and a buyer of an existing food establishment must bring it up to current building codes and food regulations [48], a cost the purchase budget must carry.

Parking and accessibility follow the seat count. The local code sets total spaces; the 2010 ADA Standards set accessible spaces by lot size, one in six of them van spaces, and require dining surfaces between 28 and 34 inches high at the accessible seats [33].

Environmental review starts with the parcel's history. SBA lenders compare the NAICS codes of a property's "current and known prior uses" with an environmentally sensitive list; a former gas station starts with a Phase I [26]. Otherwise, a loan over $250,000 begins with a questionnaire and a records search [26]. Pad sites often have fuel or auto service in their past.

Which financing paths fit a restaurant?

Six paths cover most restaurant projects. The right one depends on whether the operator owns or leases the real estate, the loan size, the location and how much of the cost is equipment.

Exhibit 27. Financing paths compared

Source: 13 CFR Part 120 (revised as of January 1, 2026); SBA SOP 50 10 8.1 and Appendix 17; 7 CFR Part 5001; USDA FY 2026 OneRD notice, 91 FR 11272. USDA resets guarantee percentages each fiscal year; confirm with the State Office.

SBA 7(a)

The 7(a) loan is the most flexible path: SBA lists real estate, working capital, equipment and changes of ownership among its uses [28]. Loans of $150,000 or less may carry an 85% guaranty and larger loans 75% [20]. Under Appendix 17, equipment, fixtures or furniture loans "should not exceed 10 years" unless the asset's IRS class life supports up to 15, leasehold improvements are limited to 10 years, and goodwill to 10 [26]. Start-ups must show coverage of 1.15 within 2 years of funding [26], and projected coverage may not include rental income from the project property [26].

SBA 504 for owner-occupied real estate

SBA describes 504 as "long-term, fixed rate financing for major fixed assets" [29]; it funds buildings, land and machinery with at least 10 years of remaining life, not working capital or inventory [29]. The SBA portion is capped at $5,000,000 for each borrower and its affiliates [20]. The business must occupy at least 51% of an existing building, or 60% of a new one [20]. Many owners hold the building in a separate company and lease it to the restaurant; SBA allows this when the lease runs at least the loan term and the rent does not exceed the loan payment plus holding costs [20]. The CDC must decide whether the property is limited or special purpose and explain why [26]. SBA's example list names gas stations and hotels but not restaurants [26], so a drive-through building is argued on its alternative uses. Each 504 project must create or retain one job opportunity per $95,000 guaranteed or meet another program objective [26].

USDA Business and Industry

The B&I program finances business acquisitions, start-ups and expansions that create or save jobs [24] in rural areas, which exclude cities and towns of more than 50,000 inhabitants and their adjacent urbanized areas [24]. For a guarantee above $1,000,000 to a new business, a feasibility study by an independent consultant is required [24]. B&I projects over $1 million that would add more than 50 jobs are ineligible where there is "not sufficient demand" to keep existing competitors at efficient capacity, unless those competitors would not be harmed [24], so the study's supply analysis does double duty. Loan funds may also pay for start-up costs and permanent working capital [24], which is where a ramp-up reserve can sit.

Conventional, equipment finance and sale-leaseback

Without a guarantee, terms come from the bank's credit policy, and the study is built to the strictest likely test. Equipment finance spreads the kitchen package over its useful life; its payment belongs in debt service. A sale-leaseback returns the owner's equity in the building but turns debt into rent, so the study reruns coverage with the new rent in occupancy cost and checks the lease term against any loan on leasehold improvements.

What will each lender ask about a restaurant?

Each capital source reads the same study through its own rules, so the questions differ by program. Exhibit 28 lists the ones that recur, with the rule behind each.

Exhibit 28. Lender questions by program

Source: SBA SOP 50 10 8.1 and Appendix 15; 13 CFR Part 120; 7 CFR Part 5001, as cited in each row.

USDA's five Cs are character, capacity, capital, collateral and conditions [24]. Under capacity the lender must ensure adequate working capital and reserves; under conditions it should submit supporting documents such as a feasibility study or market study [24]. A study organized to those headings saves a round of questions.

What must a restaurant feasibility study contain?

1. A conclusion first: feasible, feasible with stated conditions, or not feasible, with the conditions or changes spelled out.
2. The concept defined: segment and NAICS code, menu, seats, dayparts, channels and hours.
3. Program fit: size standard and affiliates, franchise Directory status, transaction category and coverage floor.
4. The trade area by daypart: drive time, workers, households and income.
5. Demand evidence: spending potential from federal data, applied to the trade area.
6. A supply survey of competing restaurants by segment, with seats, price points and pipeline.
7. A sales forecast by month, daypart and channel, built from covers, checks and turns.
8. A ramp-up path for a start-up, with the reserve that carries Year 1.
9. A cost model: costed menu, staffing schedule at local wages, occupancy, franchise fees and quotes.
10. A check of that model against Census ratios for the segment and state and, for a franchise, the FDD.
11. A development budget with sources and uses, contingency, pre-opening costs and working capital.
12. The alcohol license path: license type, food or seat test, quota cost and timing.
13. Site findings: zoning, drive-through stacking, grease interceptor, parking, accessibility, health plan review and environmental history.
14. The lease abstract tested against the loan term.
15. Coverage at the program's floor, with break-even points and downside cases for lower sales, higher food and labor cost and a slow ramp.
16. A management assessment: operating experience, multi-unit capacity and franchisor support.
17. Every source and assumption listed, with model outputs labeled as such.

How does the study run, and what drives its cost and timeline?

It runs in six steps, and scope, not the size of the restaurant, sets its cost and timeline.

1. Scoping: confirm the capital source, program, loan size and the tests that apply.
2. Document intake: concept, site plan, lease or purchase contract, budget, FDD and, for an operating restaurant, financial statements, tax returns and point-of-sale reports.
3. Field and market research: site visit, trade area, competitor survey and demand evidence.
4. Model: monthly sales and costs, base and downside cases, coverage and sales needed at the floor.
5. Review: the draft is checked against the lender's questions and revised only for new facts.
6. Final report: signed by Sarrah Allen, MAI, with written answers to the lender's follow-up questions. More about the practice is on the about page.

Scope widens with the number of units, an operating history to rebuild, a disclosure document to review, a liquor license quota or transfer, more dayparts and channels, and a 504 or USDA format. A single new franchised quick-service unit is the simplest case; a multi-unit acquisition with bars is the most involved. The fee and schedule are set at scoping.

After funding

The model becomes the yardstick for the loan's first years. Comparing monthly sales, prime cost and coverage with the forecast shows early whether a slow start is timing or demand.

For a related concept with production and taproom economics, see the brewery feasibility study. To test SBA coverage under both SOP versions, use the SBA DSCR calculator.

Download the evidence behind this page's exhibits (CSV, 770 data points with sources)