Restaurant Financial Basics: The Numbers That Decide Whether You Make It

A restaurant can be busy, well-reviewed, and dead. In the National Restaurant Association's Restaurant Operations Report, built from more than 900 restaurants, the median full-service operator reported income before taxes of 2.8% of sales; limited-service reported 4.0%. On margins that thin, a five-point error in one cost category isn't a bad quarter — it's the whole business.

Five things below: prime cost, food cost, the P&L lines that actually move, why profit and cash are different questions, and the 13-week model that ties them together. Every percentage here is a range operators commonly report — a benchmark to compare your plan against, not a target you're promised.

Prime Cost Is the Number That Matters

Prime cost is cost of goods sold plus total labor, as a percentage of total sales. The formula is (COGS + total labor) ÷ total sales. Most operators work in a band of roughly 55–65% of sales — industry guidance rather than measured data, but the band nearly every operating framework uses.

It's the majority of your controllable spend and the only large cost category you can move inside a week. Rent is set the day you sign the lease, insurance annually, debt service by the note. Prime cost is set by Tuesday's order and Thursday's schedule.

How to Calculate It

COGS is food plus beverage — everything you sell that came in on an invoice. Labor is fully loaded: wages, salaried management, payroll taxes, benefits, workers' comp.

Calculate on gross wages alone and you land several points light. How far above wages the loaded figure runs depends on your state's unemployment insurance rate — experience-rated, varying by employer — and your workers' comp classification, so pull it off your payroll register rather than using a multiplier you read somewhere.

A month with $100,000 in sales, $30,000 in food and beverage, and $32,000 in fully loaded labor:

($30,000 + $32,000) ÷ $100,000 = 62% prime cost

That leaves $38,000 for everything else, then profit. Run it on wages only, at $26,000, and it reads 56%. Same restaurant, six points of fiction.

Target Ranges by Concept

From the same NRA data: median payroll ran 36.5% of sales for full-service and 31.7% for limited-service, and full-service operators at $2M+ in annual sales reported median food and beverage cost of 31.0% — note that volume qualifier, which doesn't describe a first-year independent. Beverage COGS typically runs 15–25%, well under food's 28–35%, so a real bar program pulls the blended number down.

Notice that those two full-service medians add to 67.5% — above the 60–65% band in the table. Neither figure is wrong. The band is a planning heuristic; the medians came off real P&Ls. If your full-service model pencils at 62% prime, you are planning to run better than the median full-service operator, and you should be able to say why.

The ghost kitchen row is the one most guides miss. A delivery-only operation can post an excellent prime cost — no servers, no host, no dining room — and still not work, because platform commission isn't in prime cost. The major platforms publish tiered plans running roughly 15% to 30% of the order, plus payment processing. That comes off every third-party ticket before any ratio gets calculated. Model it as its own line — and model it above the tier rate, because sponsored listings, partly self-funded promotions, and refunded orders land on your side too. Ghost kitchen and delivery economics has the effective-take-rate math.

Why Weekly and Not Monthly

A monthly prime cost arrives four weeks after the damage, by which point you've placed four more orders against the same bad assumptions. Weekly, Tuesday's order can still respond. RestaurantOwner.com, an operator training resource, reports weekly tracking commonly moves the bottom line by two to five points — an industry claim, not a measured result. Same day, same hour, every week: a habit, not a calculation.

Food Cost, and the Number Underneath It

Food cost percentage is the headline — industry-wide, somewhere around 28–35%. It doesn't tell you what to fix.

Theoretical versus actual does. Theoretical is what your recipes say the food should have cost given what sold; actual is what left the building. The gap is the message. As a rule of thumb, variance under 2% is well controlled and over 5% signals a systemic problem rather than a bad week: over-portioning, bad yields, receiving errors, spoilage, or theft. Closing that gap is per-item work — see menu engineering and food costing for recipe cost cards, yield loss, and contribution margin.

Where Food Cost Actually Leaks

Portioning. Recipe says four ounces, line runs five, and no purchasing decision fixes that overage.

Receiving. Nobody checks the truck against the invoice, so short deliveries, substitutions, and price increases you never agreed to get paid without comment.

Spoilage. Rotation and par levels are most of it — but so is refrigeration that can't hold temperature. A walk-in drifting five degrees warm is a food cost problem that looks like a kitchen problem, and it eats product quietly for months. Price used refrigeration against what the spoilage is costing.

Labor: The Other Half of Prime

Two kinds of labor sit inside that line: the labor you schedule and the labor you're contractually stuck with. Hourly hours are a weekly decision; salaried management and benefits are a hiring decision you revisit once a year at best.

Track sales per labor hour alongside labor percentage. Percentage moves when sales move, so a slow week makes a good schedule look bad and a busy week hides a sloppy one. Sales per labor hour tells you whether the schedule fit the volume you had.

Prep equipment is a real labor lever with one caveat: a slicer or mixer only saves money if you actually cut the hours it saves from the schedule. When the math works, used prep equipment is a forgiving category — simple machines with long service lives.

Tip credit, tip pooling, service charges, minimum wage, predictive-scheduling ordinances, and unemployment insurance rates all vary by state and often by city. Several states don't permit a tip credit at all; some cities carry penalty pay for schedule changes. Check your city, not just your state, and get specifics from your state labor department.

The P&L Lines That Actually Move

Read the statement top to bottom and ask one question of each line: can I change this weekly, annually, or not at all?

The trap: small lines are the easy ones to understand, so that's where attention goes — renegotiating the linen service while prime cost runs six points hot. COGS and labor together are 55–65% of revenue; everything else combined is smaller than either one.

Repairs and maintenance is the line operators defer first, and deferring it is borrowing from next quarter at a bad rate — the compressor you didn't service becomes the one you replace mid-service, plus the product you lost. When an estimate lands, use a framework rather than a mood: when to repair and when to replace.

Occupancy Is Decided Once

Occupancy commonly runs 6–10% of gross sales, with 10% widely treated as the line where rent starts eating the business. It's the most consequential number a new operator signs, because unlike prime cost you can't fix it on a Tuesday — you fix it at renewal, or by closing. And it isn't just base rent: it's rent plus CAM, property taxes, and building insurance. A comfortable base rate can land outside the band once pass-throughs are added.

Equipment on the P&L

Buy a range outright and that's capital expenditure — it lands on the balance sheet as an asset, not in your expense lines. What shows up monthly is depreciation (non-cash), repairs and maintenance (very much cash), and financing payments if any. That's why "I bought used and saved $40,000" shows up dramatically in cash flow and almost invisibly on the P&L.

The restaurant equipment checklist is the source for what a full kitchen costs new versus used — the figure that sets the opening cash position the 13-week model starts from — and used cooking equipment shows what the market charges. Finance it and the payment splits: interest is an expense, principal isn't, and the whole payment still leaves the account. See how equipment financing works.

Profit and Cash Are Not the Same Thing

A restaurant can post a profitable month and miss payroll. That's not an accounting error and it isn't rare. The P&L records revenue and expenses in the period they belong to; cash moves on a different schedule:

• Card settlement lags — Friday's sales land Monday or Tuesday, minus processing, and delivery payouts run on their own cycle entirely.
• Vendor terms of 7 to 30 days — 45 or 60 on some accounts — mean week one's food gets paid for in week four or later.
• Biweekly payroll lands three times in some months. Your P&L smooths it; your bank account doesn't.
• Quarterly payroll tax and sales tax remittance are large, lumpy, and not optional.
• An annual insurance premium can arrive as a single bill.
• A deposit on a walk-in is a full cash outflow that shows up on the P&L as nothing.

Sales tax deserves its own warning: what you collect is not your money. Rate, base (prepared food is often taxed differently than groceries), and filing frequency vary by state and often by city, and frequency changes as volume grows. Some cities add a separate meals tax. Ask your state department of revenue for your filing frequency, put remittance on the disbursements side of the model, and never spend it. Same for license and permit renewals — covered in the restaurant permits and licenses checklist.

Depreciation, Principal, and the Two Directions the Gap Runs

Depreciation reduces profit without touching cash. You bought the equipment once. Accounting spreads that cost across the years you'll use it, so a slice lands on the P&L monthly as an expense. No money moves, and profit looks lower than the bank account feels.

Loan principal consumes cash without touching profit. The interest portion of a payment is an expense; the principal portion is repayment of a balance, not an expense at all. The whole payment still leaves the account. Those two produce the classic profitable-but-broke restaurant.

How Much Cushion

Planning guidance for a new operation is a working capital reserve of three to six months of fixed operating expenses — rent, base payroll, insurance, debt service, and software, the costs that arrive whether or not guests do; established operations are commonly held to two or three. Both are rules of thumb with wide variance — a food truck's expenses and a 90-seat dining room's are different animals. What isn't variable: the reserve stays separate from your build-out budget, because money spent on equipment isn't available for payroll in month three. For sizing the total, see how to estimate restaurant startup costs.

None of This Works on Books That Are Three Weeks Old

Weekly prime cost requires last week's invoices entered. A 13-week forecast requires the bank reconciled and payroll categorized. Books three weeks behind aren't a management tool — they're a history project.

Most first-time operators do their own books for about four months and then stop, usually around the first quarterly deadline. The work doesn't stop; it accumulates. Three honest options:

Do it yourself, with discipline. Two to four hours a week — invoice entry, categorization, bank reconciliation — plus a longer session at month end. Doable, but it means doing it on the weeks service went badly, which is exactly when you won't want to.

Hire a bookkeeper. Someone else handles entry, reconciliation, and the close; you get statements you can act on.

Run a hybrid. You capture daily sales and enter invoices; a bookkeeper does reconciliation and the close. Most operators land here.

If you go looking: full disclosure, InsightTrack Bookkeeping is owned by the same person who owns this site — no affiliate deal, no commission, just common ownership, and you should know that before you click. It's a general small-business practice on QuickBooks: monthly reconciliation, categorization, reporting, cash flow review, flat monthly pricing. It's not a restaurant specialist and this page won't pretend otherwise. A general bookkeeper who keeps your books current beats no bookkeeper by a wide margin; if you want restaurant-specific expertise on top, talk to a hospitality CPA too.

The 13-Week Cash Flow Model

A 13-week cash flow forecast is a rolling weekly projection of cash in and cash out across one quarter. Every week: the week that just closed gets actuals in place of forecast, a new week 13 is added at the far end, and each week's closing balance becomes the next week's opening balance.

Why 13 and not 12 or 26? Thirteen weeks is one quarter — long enough to contain a full cash cycle (monthly billing, biweekly payroll landing three times in some months, quarterly tax remittance, 30- to 60-day vendor terms), short enough that the numbers stay grounded in what you know rather than invented.

What the Sheet Looks Like

Three row blocks, thirteen columns, one column per week.

Opening cash. One row. Week 1 is today's actual bank balance.

Receipts. Card settlements dated when they land, not when you rang the sale. Cash sales. Catering deposits. Delivery platform payouts on their own row, because they arrive on a different schedule and blending them hides the timing.

Disbursements. Food and beverage vendors, payroll, payroll taxes, rent and occupancy, utilities, insurance, debt service, sales tax remittance, license and permit renewals, equipment purchases. Give equipment its own row — a single used commercial kitchen equipment purchase can be the largest disbursement in a quarter, and it's the one most often forecast as a deposit and paid as a balance.

Closing cash. Opening plus receipts minus disbursements, carried into next week's opening. The row of closing balances across thirteen weeks is the entire point. That's the whole model — a spreadsheet you can build in an afternoon.

The Weekly Rhythm Is the Tool

Same hour, every week. Replace last week's forecast with actuals. Give every variance over your threshold a one-line reason — not analysis, just "produce order $800 over, three-day event." Add the new week 13. Read the closing balance row.

The spreadsheet is trivial; the habit is the thing. The model's job is to show you a shortfall six weeks out, while you still have options — cut an order, delay a hire, ask a vendor about terms, move an equipment purchase a month — instead of the Friday it arrives, when the only choice left is which bill not to pay.

If you'd rather not build it from scratch, the 13-week cash flow template below has the row structure laid out, a restaurant-specific disbursement list including the lines operators forget, and the formulas wired so closing balances carry forward. Put in your opening balance and start forecasting.

What to Ask Your CPA or Bookkeeper

This page explains how the numbers work. What to do about yours is a conversation with someone who knows your situation, your state, and your entity. Go in with questions:

• Should I be on cash or accrual accounting, and why, for a business like mine?
• How should COGS be structured in my chart of accounts so prime cost falls out of it automatically instead of needing a manual calculation every week?
• What's my sales tax filing frequency, what does my state tax on a restaurant ticket, and at what volume does the frequency change?
• How should equipment purchases be recorded, and what does that do to my taxable income? (Section 179 and bonus depreciation are the terms to ask about — don't decide it from an article, including this one.)
• What entity structure makes sense here, and how should I be paid out of it? The legal structure guide lays out what the options are so you walk in with the question already framed.
• What does my state do with tip credit and tip pooling, and what's my actual unemployment insurance rate?

Every one has a cash consequence and depends on facts about you. The ratios here are how the game is scored; what to do on your field is a conversation.

Frequently Asked Questions

What is prime cost in a restaurant?

Prime cost is cost of goods sold — food and beverage — plus total fully loaded labor, divided by total sales. Fully loaded labor means wages, salaried management, payroll taxes, benefits, and workers' comp, not just hourly wages. Most operators work in a band of roughly 55% to 65% of sales, quick service lower and full service higher.

What percentage should food cost be in a restaurant?

Industry-wide it commonly lands in the 28% to 35% range: quick service and fast casual around 25% to 30%, casual dining 28% to 35%, fine dining 30% to 40%. Beverage cost of goods runs lower, roughly 15% to 25%. These are ranges operators report, not targets — and the more useful measurement is the gap between your theoretical food cost and your actual one.

Why is my restaurant profitable but has no cash in the bank?

Profit and cash run on different clocks. Card settlements lag sales by days, vendors get paid on terms running anywhere from 7 to 60 days, biweekly payroll lands three times in some months, and quarterly tax remittances are lumpy. Depreciation also reduces profit without money moving, while loan principal drains cash without appearing as an expense — so a profitable month can still miss payroll.

What is a 13-week cash flow forecast and why do restaurants use one?

It's a rolling weekly projection of cash receipts and disbursements across one quarter, refreshed every week — actuals replace the forecast for the closed week, a new week 13 is added at the far end, and each closing balance becomes the next opening balance. Thirteen weeks contains a full cash cycle: biweekly payroll, quarterly tax remittance, 30- to 60-day vendor terms. It surfaces a shortfall weeks before it arrives.

How much cash should a new restaurant have in reserve?

Common planning guidance for a new operation is three to six months of fixed operating expenses in working capital — rent, base payroll, insurance, debt service, and software — with established operations often holding two to three. Treat it as a rule of thumb — the right number depends on your concept, fixed costs, and seasonality. What matters more is that the reserve stays separate from your build-out budget, because money spent on equipment isn't available for payroll in month three.

Should I do my own restaurant bookkeeping?

Plenty of operators do, and it's a legitimate choice. Realistically it costs two to four hours a week plus a longer session at month end, and it only works if you keep to it during the weeks service went badly. The common failure isn't incompetence, it's falling three weeks behind, at which point the books stop being a management tool. Put it on the calendar as a fixed appointment; if that keeps losing to service, hire it out or split it.