Insights / Reference
Restaurant and hospitality KPIs
Every metric below is defined the way operators actually use it: the calculation, and the trap that most often makes it wrong. At the end, a worked month in which the headline number improves and the business makes less money.
Published by Back of House Accounting · Principal: Douglas Carpenter, CPA, CFA · Last updated 25 September 2026
Restaurant
Food, beverage and labor
| Term | Calculation | What it tells you, and where it goes wrong |
|---|---|---|
| Prime cost | (Cost of sales + total labor) ÷ net sales | The single most useful restaurant number, because it covers the two costs that move weekly. Goes wrong when payroll taxes and benefits are excluded from labor — that understates it by two to four points. |
| Theoretical food cost | Recipe cost × items sold | What food cost should have been. The gap against actual is waste, theft, over-portioning or mispricing. Useless unless recipes are current — a costing from eighteen months ago measures nothing. |
| Average check | Net sales ÷ covers | Separates traffic growth from price growth. Goes wrong when covers are counted as tickets — a two-top and a party of eight are not the same cover count. |
| Sales per labor hour | Net sales ÷ hours worked | Productivity independent of wage rate, which makes it comparable across units in different wage markets. |
| Comp, void and discount rate | Comps + voids + discounts ÷ gross sales | Where restaurant losses hide. Worth watching by server and by daypart, not as a single blended figure. |
| Four-wall contribution | Unit revenue − all unit-level costs, before group overhead | The honest measure of whether a location earns its place. Goes wrong when group overhead is allocated in — that measures the allocation method, not the restaurant. |
Lodging
Rooms, profit and conversion
| Term | Calculation | What it tells you, and where it goes wrong |
|---|---|---|
| ADR | Rooms revenue ÷ rooms sold | Rate achievement. Goes wrong when complimentary and house-use rooms are left in the denominator, which drags the rate down artificially. |
| RevPAR | Rooms revenue ÷ rooms available | Combines rate and occupancy into one figure, which is why it is the standard comparison. It says nothing about profit — RevPAR can rise while GOP falls. |
| TRevPAR | Total revenue ÷ rooms available | Captures food, beverage and other departments. The better measure for full-service and resort properties where rooms are under half of revenue. |
| GOP | Total revenue − departmental and undistributed expenses | Profit before fixed charges and ownership costs. The number a management company is generally measured on. |
| Flow-through | Change in GOP ÷ change in total revenue | How much of each incremental revenue dollar reached profit. The most diagnostic lodging metric and the least used. Below roughly 40% on a revenue gain means cost grew with volume in a way it should not have. |
| CPOR | Rooms department expense ÷ rooms sold | Cost per occupied room. Rising CPOR alongside rising occupancy usually points at labor scheduling rather than at demand. Which accounts make up rooms department expense is set out in the hotel chart of accounts. |
| USALI | — | The Uniform System of Accounts for the Lodging Industry: the standard presentation brands, lenders and asset managers expect. Reporting outside it triggers rework at every review. |
Payroll, tax and calendar
Terms that cause the most trouble
| Term | What it means and why it matters |
|---|---|
| Tip credit | The amount of an employee's tips an employer may count toward the minimum wage obligation. Rules vary by state and several states allow none at all. Applied incorrectly across a multi-state group it becomes a wage claim. |
| FICA tip credit | A federal income tax credit for the employer share of Social Security and Medicare paid on reported tips above the minimum wage threshold. Frequently unclaimed — it is real money left on the table every year. |
| Form 8027 | The annual return for large food and beverage establishments reporting tip income and gross receipts. Triggered by employee count and tipping customarily occurring; missing it invites examination. |
| Service charge vs gratuity | A mandatory service charge is wages, not a tip — it is subject to payroll tax and generally cannot be counted toward tip credit. Getting this backwards is one of the most common and most expensive hospitality payroll errors. |
| 4-4-5 calendar | A fiscal structure of 4, 4 and 5-week periods per quarter, so each quarter contains thirteen weeks and comparisons hold. Requires a stated period calendar or reporting becomes ambiguous. |
| Deferred revenue | Cash collected before the service is delivered — event deposits above all. It is a liability until the event occurs. Recognizing it on receipt overstates profit now and creates a hole later. |
| Percentage rent | Additional rent owed once sales exceed a stated breakpoint. Belongs in occupancy, accrues as sales are earned, and should never be left to a year-end true-up. |
Worked example
A month where prime cost improved and profit fell 36%
Prime cost is the number most operators watch, and a falling prime cost percentage is normally read as a win. It is a ratio, and a ratio can be improved by shrinking the thing underneath it. Here is a single 100-seat restaurant across two months. The only decisions taken were an 8% price increase, after which covers fell 18%.
| Before | After | Move | |
|---|---|---|---|
| Covers | 4,000 | 3,280 | −18.0% |
| Average check | $42.00 | $45.36 | +8.0% |
| Net sales | $168,000 | $148,781 | −11.4% |
| Cost of sales | $50,400 | $41,328 | |
| Food cost % | 30.00% | 27.78% | better by 2.2 pts |
| Total labor | $58,800 | $52,536 | |
| Labor % | 35.00% | 35.31% | worse by 0.3 pts |
| Prime cost % | 65.00% | 63.09% | better by 1.9 pts |
| Fixed costs | $48,000 | $48,000 | unchanged |
| Profit | $10,800 | $6,917 | −36.0% |
Cost of sales modelled at $12.60 per cover. Labor modelled as $24,000 fixed plus $8.70 per cover. Both months reconcile: net sales − prime cost − fixed costs = profit.
Read only the percentages and this was a good month. Prime cost fell almost two points and food cost percentage fell more than two. A prime cost of 63% would pass most reviews without comment.
The business made $3,883 less. Three things did it, and each one is a reason the percentage misled:
- Cost of sales is close to purely variable. Losing 18% of covers removed 18% of food cost automatically. That improves the ratio without anyone controlling anything.
- Labor is not purely variable. Salaried management, openers and closers do not scale with covers, so labor cost percentage got worse on falling sales even though labor dollars fell. This is the tell, and it is the one line most reports do not separate.
- Fixed costs did not move at all. Rent, insurance and administration were the same $48,000 against $19,219 less sales. Percentages do not pay rent; dollars do.
The diagnostic that catches this in one line is contribution dollars per period, not any percentage: net sales less variable costs, tracked in dollars. It fell from $82,800 to $75,917 and told the true story immediately while prime cost was still reporting an improvement.
How we read them
The order to look at these in
This is our own sequence rather than an industry standard, and it exists because reading these numbers in the wrong order is how an operator reaches the right conclusion two months late.
- Sales, split into covers and average check. Every ratio below sits on this denominator, so a move here explains most surprises further down. A ratio that improves on falling covers is the case above.
- Contribution dollars. Sales less variable cost, in dollars. Immune to the trap in the worked example.
- Prime cost, fully burdened. Only after the first two, and only with payroll taxes and benefits inside labor.
- Labor split fixed against variable. The single most useful split in restaurant reporting and the one most rarely presented.
- Comps, voids and discounts by server and daypart. Never as a blended figure.
- Four-wall contribution by unit, before any group overhead allocation.
A monthly close that lands on the 25th cannot support any of this, which is why we commit to a closed profit and loss statement and balance sheet by day 10. If you want these numbers produced and read every period rather than calculated once, that is CFO and controller work.
Questions
Questions operators ask about these numbers
What are KPIs in a restaurant?
Key performance indicators are the small set of measures that describe whether a restaurant is working: sales split into covers and average check, cost of sales, labor, prime cost, comps and discounts, and four-wall contribution by unit. The useful ones are the ones tied to a decision. A measure nobody acts on is a report, not a KPI.
What KPIs should a restaurant track weekly rather than monthly?
Cost of sales, labor hours and dollars, covers, average check and comps. These are the figures that still move during the period, so a weekly flash lets you act inside the month. Fixed costs, rent and occupancy do not need weekly attention because nothing you do that week changes them.
What is a good prime cost percentage for a restaurant?
Prime cost is judged against the same restaurant's own trend and its service model before any external figure, because a quick-service operation and a full-service one with a large kitchen brigade are not comparable. The worked example above is the reason: a prime cost percentage can improve while the business earns less, so the level matters far less than what moved to get there.
Why did my food cost percentage improve when business got worse?
Because cost of sales is close to purely variable. If covers fall, food cost falls with them automatically and the ratio improves without anything being controlled. Check contribution dollars and covers before treating a falling food cost percentage as a result.
What is the difference between RevPAR and GOPPAR?
RevPAR is rooms revenue divided by rooms available, so it combines rate and occupancy but says nothing about cost. GOPPAR is gross operating profit per available room, which carries departmental and undistributed expense. RevPAR can rise while GOPPAR falls, which is why an owner and a revenue manager can look at the same month and disagree.
Is a service charge a tip for KPI and payroll purposes?
No. A mandatory service charge is wages rather than a tip, which changes payroll tax treatment and generally cannot be counted toward a tip credit. It also belongs in revenue, so it moves average check and every ratio built on sales. The distinction is covered in full in service charge versus tip.
Restaurant profitability guide
More on restaurant profitability
Service
Restaurant accounting services
Prime cost by unit, period close and multi-unit reporting.
Guide
Prime cost
What prime cost is, how to calculate it correctly, and why a monthly figure arrives too late.
Guide
Food cost percentage
The formula, theoretical against actual, and six treatments that move the number for no operational reason.
Guide
Labor cost percentage
A fully burdened labor figure, fixed versus variable, and the denominator traps.
Guide
Restaurant cash flow
Why profitable restaurants run short on cash, and the 13-week forecast that shows it coming.
Guide
Restaurant P&L statement
The P&L format line by line, a worked period close, and a blank template to copy.
Guide
How restaurant accounting works
The full cycle: period calendars, POS settlement, inventory, prime cost, tips and the close.