How to manage kitchen staff: the NUMBERS that separate the traditional method from the Masterestaurant method

What should kitchen payroll cost as a share of food sales?

Between 14% and 18% of food sales, and that range gets settled with a schedule, not with charisma.

The figure comes from breaking down the total labor cost the National Restaurant Association (2025) measured at 36,5% of sales for full service and 31,7% for limited service during 2024: the kitchen takes roughly half, the dining room and administration the rest. The Bogotá group that opened this analysis was running 24,8% in back of house alone, six points above the ceiling, with four locations billing evenly. When you measure that percentage week by week instead of month by month, the deviation shows up on Tuesday and you can still fix it by Thursday; measured over thirty days, it has already turned into an accounting loss and all that remains is explaining it to the board. Two extra cooks between 15:00 and 17:00, every single day, cost around 1.400 dollars a month in a mid-sized location — more than the raise the owner refused to give and that eventually triggered the resignation.

The cost is not in the wage: it is in the badly placed hour

That is the finding that stings most when you cross the schedule against ticket volume by time band: the dead afternoon stretch concentrates the waste, while the cold station calls for backup at 21:10 because nobody shifted the turn ninety minutes. The National Restaurant Association (2025) documented the gap plainly: PROFITABLE full-service restaurants closed 2024 at 34,2% labor cost and money-losing ones at 42,9%. Eight and seven tenths of a point do not come from different hourly rates; they come from who is standing in the kitchen when not a single order is coming in. Replacing an hourly worker costs 2.305 dollars in hard separation, recruiting and training expenses, according to Black Box Intelligence (State of Restaurant Workforce 2024), while the Cornell Center for Hospitality Research put the average cost of losing a frontline employee at 5.864 dollars. The distance between both numbers is exactly what accounting never records: the weeks that position runs at 60%, the overtime the rest of the line absorbs, the waste a new cook generates working without a recipe spec.

Turnover gets paid twice and only one payment reaches the books

With back-of-house annual turnover under 60% —my working threshold— a twelve-person kitchen replaces seven slots a year and burns about 41.000 dollars. Push it to 110%, which is ordinary in this industry, and the bill crosses 77.000. There sits the raise nobody wanted to give. The operating minimum is 18 hours of structured training for every person who joins, spread across the first two weeks and logged with signature, station and date. Nobody in that Bogotá group knew how many hours each hire had received, and that blindness is what produces invisible retraining: six months later somebody explains the same cold-station protocol again, now at peak hour with the line running. Turn it around, which is where the number really bites. Skip the training and save those 18 hours at local rates, and you pocket maybe 90 dollars per hire; if that person quits in month three, you just traded 90 dollars for the 2.305 Black Box Intelligence reports.

Eighteen training hours per hire, counted and signed

No staffing decision carries an asymmetry this brutal or this easy to avoid. Most restaurant management courses teach you to cost the recipe and build the roster, then let go of your hand right at the point that decides the check: the pass and the table. When the dining room does not know what went wrong on the line, the server improvises an apology and average check drops; when the kitchen does not know what the server promised, the plate comes out flawless and the guest leaves annoyed anyway. Diego F. Parra has spent twenty years walking into kitchens with the P&L in one hand and the pass timer in the other, and the Masterestaurant conclusion is uncomfortable for anyone managing on instinct: most of the labor cost is not lost in wages, it is lost in badly assigned hours and in the handoff nobody ever wrote down. Track returns by station and by shift for two weeks, and the problem names itself.

The skills gap does not close by hiring better: it closes by training differently

Kitchen vacancies have been the hardest to fill in the entire industry since 2021, so competing for the same scarce line cook is a losing strategy before it starts. SHRM cost-per-hire benchmarks (2025) place hourly and frontline hiring between 1.000 and 2.500 dollars, with the general non-executive average at 5.475 dollars. A restaurant that can take a prep hand to line cook in four weeks stops paying that lottery and hires attitude instead of résumé. I got this wrong for years: I treated the recipe spec as a costing instrument, when it is above all a TRAINING instrument, because it turns the chef's judgment into something transferable to somebody who walked in on Monday. No benchmark exists so you can copy an average: it exists so you can ask a question. In a SMALL location, a single kitchen with fewer than eight people, the 14-18% usually breaks because the owner-chef never counts himself in payroll; book yourself at market rate and the real number appears.

How to read these numbers in YOUR operation?

In a MID-SIZED location, with two shifts and weekend rotation, the lever is almost always the 15:00 to 17:00 band and the 1.400 dollars a month evaporating there.

In a GROUP of three or more, the issue stops being the schedule and becomes comparability: if each kitchen defines a training hour differently, you do not have four data points, you have four opinions. Start by unifying the definition and argue about the percentage afterwards. The labor cost percentages come from the National Restaurant Association (2025) on 2024 results and from the U.S. Bureau of Labor Statistics, which places industry labor cost between 25% and 35% of revenue; the turnover figures come from Black Box Intelligence (2.305 dollars per hourly replacement), from the Cornell Center for Hospitality Research (5.864 dollars per frontline employee, a study with years on it already) and from the 2025 SHRM benchmarks.

Where these benchmarks come from and how far they reach?

All of them are U.S. market data and medians, not averages: half the sample sits above the line.

Moving them to Latin America requires adjusting for hourly rates and for payroll burden, which in several countries adds between 40% and 55% on top of base salary. Use them as a directional thermometer, never as an imported target. The cost is not in the wage, it sits in the badly placed hour. Two extra cooks between 15:00 and 17:00, every single day, add up to roughly 1,400 dollars a month in a mid-size unit: more than the raise the owner refused to give, the one that triggered the resignation. Turnover gets paid twice, and only one of those payments shows up in the books. Replacement costs recruiting and training, true, but it also costs the weeks when that position runs at 60% while the rest of the line covers with overtime.

Where kitchen management actually breaks?

The industry skills gap does not close through better hiring, it closes through different training.

Kitchen roles have been the hardest to fill in the industry since 2021, so the restaurant that can take a prep cook to line cook in four weeks stops competing for talent that does not exist. Kitchens are managed from the pass, not from the office. The number that moves the result is not recipe cost, it is how many minutes pass between the server firing the ticket and the plate leaving hot, because table turns, upselling and tips all hang off that figure. A poorly briefed server sends the kitchen's work back. When the floor does not know the daily special, its allergens and its margin, the kitchen cooks well while the business sells badly, which is the most expensive way to be right.