Restaurant & Hospitality Finance Benchmarks 2026

Labor cost from the Bureau of Labor Statistics, menu prices from the CPI, and the cash cycle restaurants actually run.

Restaurant and hospitality financial management

Key Takeaways

  • •Food services and drinking places employed 12,390,900 people in September 2026, at an average hourly wage of $22.13 (BLS).
  • •Food away from home prices rose 3.5% over the 12 months to September 2026 (BLS CPI, via FRED).
  • •Restaurants collect almost immediately: the median days sales outstanding across public restaurant filers was 5.8 days (SEC XBRL frames, n=23).

Restaurant labor, September 2026

12,390,900
Employment, food services and drinking places
BLS Industry at a Glance: Food services and drinking places (NAICS 722), Nov 10, 2026
$22.13
Average hourly earnings, all employees
BLS Industry at a Glance: Food services and drinking places (NAICS 722), Nov 10, 2026
25.0
Average weekly hours, all employees
BLS Industry at a Glance: Food services and drinking places (NAICS 722), Nov 10, 2026
$20.39
Average hourly earnings, production and nonsupervisory
BLS Industry at a Glance: Food services and drinking places (NAICS 722), Nov 10, 2026

Labor is the cost, and here is the data behind it

Restaurant labor is the largest expense and the one that scales with the schedule rather than with sales. The Bureau of Labor Statistics reports 12,390,900 people employed in food services and drinking places in September 2026, with average hourly earnings of $22.13 across all employees and $20.39 for production and nonsupervisory workers, on an average 25.0-hour week (BLS Industry at a Glance: Food services and drinking places, NAICS 722, data extracted November 10, 2026).

Multiply that out and you get the number that drives every labor decision: $22.13 an hour, across a 25-hour week, across 52 weeks is about $28,800 a year of wages per employee before payroll tax and benefits. At a 25% labor cost target, a $1M restaurant carries roughly $250,000 of labor — about nine full-time-equivalent positions, or the same money spent on fifteen part-time schedules.

That is the conversion we use in an audit. Percentages are abstract until they turn into shifts, and shifts are where the money actually goes.

Turnover: use the quits rate, not folklore

This page used to carry a turnover figure of 70% annually for front-of-house staff. It was folklore, and it is gone. The sourced version is the Bureau of Labor Statistics Job Openings and Labor Turnover Survey: the quits rate for accommodation and food services averaged 4.2% per month over the twelve months to September 2026, essentially unchanged from 4.1% over the prior twelve months (FRED series JTS7200QUR, from BLS JOLTS). For comparison, the all-industry quits rate was 2.1% in September 2026 (BLS JOLTS news release, November 3, 2026).

A monthly quits rate near 4% compounds into heavy replacement hiring in a business where the ramp to productivity is weeks rather than days. Two things we do with that number. We treat quits as a leading indicator of labor cost, because an unscheduled absence costs a shift and an unstaffed table costs a cover. And in states with tipped minimum wages, we treat turnover and the tip structure as one financial question rather than two: the tip credit only makes sense if the servers it applies to stay long enough for the operator to recover the training and recruitment cost.

What restaurants look like in the filings

Public restaurant filers are a small but informative group, and they show the cash mechanics every operator already knows. Among companies with a calendar fiscal year ending December 31, 2024 and SIC 5812 (eating places), the median days sales outstanding was 5.8 days, computed from the 23 companies that tagged both receivables and revenue.

Two conclusions follow. Restaurants collect faster than any other sector we measured, so a receivable problem is almost always a card-processor or a banquet-contract problem rather than a collections problem. And the working capital in a restaurant is not receivables at all. It is the decision about how hard to push food and beverage vendors.

We do not publish a restaurant cash conversion cycle. A cycle needs inventory, payables and cost of revenue as well as receivables, and only three restaurant filers in this sample tag all four — a median of three companies would tell you nothing. The receivables median above is the part of the picture that holds, and the vendor-terms decision is the part you control.

Median revenue growth for the same group was 2.5% (n=22). If you are growing more slowly than the public restaurant median, price is the fastest lever you have, and menu mix is usually what funds it without losing volume.

Food cost, prime cost and profit: our operating targets

We no longer present food cost and prime cost percentages as benchmarks, because we could not source a national benchmark for them. What follows are the targets we hold client restaurants to. They are targets, and we label them that way.

Food cost: 25-35% of food sales. The top of the band belongs to fine dining, where the ingredient cost is part of the product; the bottom belongs to quick service, where the menu is engineered for speed and volume. Casual dining usually sits in the middle.

Prime cost (food plus labor): under 65% of revenue for full-service concepts, under 60% for quick service. Above 65%, marketing cannot rescue the restaurant, because the money is not there.

Net profit: 3-8% for a well-run operation. That sounds thin until you multiply it by the revenue base and by the asset intensity of the location.

The test we run first: if prime cost is over 65%, the business is a staffing and purchasing problem, not a marketing problem. If it is under 60%, there is capital for a second location.

How to think about food cost

A food cost percentage on its own is a trap, because it is a ratio and both sides of it move. A restaurant with 32% food cost on $2M of food sales generates $1.36M to cover everything else. A restaurant with 28% food cost on $1.8M generates $1.30M — better percentage, less money, and a worse year. Volume, price and purchasing have to be managed together, which is why we look at gross profit dollars by menu category rather than at the blended percentage.

The second trap is mix. A blended food cost percentage that improved because the bar sold more liquor is not an operating improvement in the kitchen. We ask for food cost broken out by menu category, by day part, and by station before we accept any improvement claim.

Beverage is the highest-return department

Our rule of thumb, to be replaced by your own measured numbers: liquor typically costs 15-20% of drink sales, beer 20-30%, and wine 30-45%. A well-designed bar program therefore earns a materially better margin than the dining room does, and the highest-margin hours on a Saturday are not the dinner rush.

The measurement that matters is contribution per labor hour by day part. A restaurant that sells more drinks at 4pm on a Tuesday and fewer at 8pm on a Saturday has a scheduling problem, not a beverage problem.

Menu engineering without the folklore

Menu engineering is a financial method, not a template. We classify items by popularity and contribution margin: stars (high popularity, high margin) get protected placement and availability; puzzles (low popularity, high margin) need marketing and repositioning; workhorses (high popularity, low margin) may earn their place as traffic drivers, but only if you know what they are worth; dogs (low popularity, low margin) come off the menu.

What we will not do is quote the widely repeated statistic that the bottom 20% of items produce only 10-15% of sales. It circulates widely and we could not source it, so instead we give you the method: pull contribution margin and units sold for every item, sort, and look at the bottom quartile yourself. It takes an afternoon and the answer is specific to your restaurant.

Menu mix should be reviewed quarterly. Removing or repositioning items simplifies operations as well as improving margin, and simplification is worth money in a kitchen that is short-staffed at dinner.

Technology: judge it on three numbers

Restaurant technology decisions are usually argued on features. We argue them on three numbers, and if a project moves none of them we do not recommend it.

Average ticket time. Lower is better, and it is measurable before and after.
Order error rate. Every error is a remade plate, a comp, and a server who now has a problem at the pass.
Manager hours spent on inventory counts. This is the number nobody tracks, and it is usually the largest hidden cost of a POS or inventory system.

On that basis: point-of-sale modernization is justified when ticket time and order errors are both material; kitchen display systems pay for themselves when ticket time and manager count hours both move; inventory systems pay for themselves when waste and count hours both fall. A project that only produces better reports is a reporting project, and it should be priced like one.

On delivery: third-party platforms charge a commission on every order, and the economics swing on average ticket size, distance and volume. Our rule is to model the contribution per order after commission, not the order count.

Cash flow, working capital and reserves

Restaurants have unusual cash patterns — weekly cycles between slow weekdays and full weekends, payroll-driven monthly cycles, and strong seasonality — which means a cash reserve has to cover three different cycles at once.

Our operating standard: 4 to 6 weeks of cash reserves for a single-unit restaurant, longer if there is seasonality or a lease reset coming. We set the reserve by weeks because it maps to payroll periods, which is the granularity at which a restaurant actually runs out of money.

On capital reserves, our standard is 3-5% of revenue set aside for equipment replacement, and we prefer a scheduled reserve to a credit decision made in a cash crunch. Kitchen equipment ages on its own schedule regardless of what the business is doing that year.

Food safety compliance is a cost centre

Food safety obligations — training certification, temperature control and sanitation, monitoring and documentation, insurance, and the record-keeping that proves all of it — run through every shift and every inspection. We treat them as operating cost, not overhead to be trimmed, and we review insurance coverage against the specific operations rather than accepting whatever the last broker sold.

The failure modes we plan against are not the fines; they are the closure, the reputational damage, and the insurance exclusions that arrive afterward. A single serious incident can end a restaurant, and the number that ends it is rarely a penalty.

Hospitality finance: the hotel version

Hotels add real estate, 24-hour staffing and seasonality to the restaurant cost structure, and they measure performance differently: RevPAR (revenue per available room), GOP (gross operating profit) and GOPPAR (GOP per available room). Those are definitions, and definitions are not benchmarks — the level of RevPAR that works depends on market, class and local supply, and we do not have a sourced benchmark for it.

What is true across hotel properties: labor scheduling follows occupancy rather than a fixed roster, and the peak-to-trough staffing swing is where the margin is made or lost. Meeting and event revenue usually carries a materially better margin than rooms, and rooms revenue at budget price points often barely covers housekeeping. We ask for revenue mix by department before we ask for occupancy.

Sources

Labor. BLS Industry at a Glance: Food services and drinking places (NAICS 722) — employment, average weekly hours and average hourly earnings, extracted November 10, 2026, reference month September 2026 (preliminary).

Menu prices. Consumer Price Index for All Urban Consumers: Food away from home — the 3.5% twelve-month rise to September 2026.

Turnover. JTS7200QUR — the monthly quits rate for accommodation and food services, seasonally adjusted, from BLS Job Openings and Labor Turnover.

Wages across the economy. BLS Employment Cost Index — private industry wage growth of 3.1% over the twelve months to September 2026.

The receivables median. Computed from SEC XBRL data over calendar FY2024, same method as the working-capital report: ending receivables divided by revenue, times 365. The script that does it is `scripts/xbrl-benchmarks.py`.

What is ours rather than published. The sample selection, the median rather than the mean, and the choice to report a negative cash conversion cycle as a feature rather than a warning are our analysis. BLS publishes the labor and price data; it does not publish a restaurant's cash cycle.

Strengthen your hospitality finance

We will rebuild your prime cost, labor schedule and cash reserve from the numbers above, and tell you which lever moves first.

Start a conversation

Frequently Asked Questions

What is a healthy prime cost for restaurants?

Under 65% of revenue for full-service concepts and under 60% for quick service is our operating standard. Above 65% the restaurant is a labor and purchasing problem rather than a marketing problem. We publish these as our targets, not as a market benchmark, because we could not source a national one.

How can restaurants reduce food costs?

Menu engineering on contribution margin rather than popularity alone, portion control, purchasing terms, and a waste log that gets reviewed weekly. Reduce blended food cost and it is usually mix and waste moving, not pricing.

What causes restaurant labor cost increases?

Wage rates, overtime from thin scheduling, turnover and the training and productivity loss that follows it, and benefit costs. Wage inflation is running ahead of the operator's menu pricing: private industry wages rose 3.1% over the 12 months to September 2026 (BLS Employment Cost Index), while food away from home prices rose 3.5% over the same period (BLS CPI via FRED).

How do hospitality businesses manage seasonal fluctuations?

Flexible part-time and seasonal staffing, variable lease or management fees where the landlord will allow it, cross-training to redeploy staff, and building cash reserves during the peak rather than funding the trough with debt.

What technology investments provide the best ROI for restaurants?

We evaluate against three numbers: average ticket time, order error rate, and manager hours spent on inventory counts. A system that improves only the reporting is a reporting investment and should be priced like one. We do not publish vendor price ranges or percentage savings, because they change and we could not source a current, reproducible figure.

How does menu engineering improve restaurant profitability?

It sorts the menu into stars, puzzles, workhorses and dogs by popularity and contribution margin, so that high-margin items get the placement and the availability and low-margin items get removed or repositioned. Review it quarterly. The specific statistics on how much revenue sits in the bottom of the menu are the numbers we would most want to see sourced, and we have not found a method we can reproduce.

What pricing strategies work best for restaurants?

Cost-plus for commodity items where you can defend the price, value-based where the item has a differentiated benefit, competitive pricing where you are transparent that you match, and psychological pricing for perception. Promotional pricing needs cannibalization analysis — happy hour frequently buys price-sensitive customers who would not otherwise visit, at a discount to the ones who would.

How should restaurants plan for capital expenditures?

Our standard is 3-5% of revenue for equipment and dining room replacement, built as a scheduled reserve rather than a credit decision in a crunch. Kitchen equipment ages on its own schedule whether or not the business is doing well.

Dan Settel
Dan Settel
Co-Founder, Eagle Rock CFO