What Turnover Costs a Restaurant Group Per Position

The problem: Turnover is treated as a fact of hospitality rather than as a purchase, so nobody has ever put a price on filling a position and nobody can compare that price to the cost of keeping someone.

The solution: Cost a departure the way you cost a food order, per position, including the trainer's hours and the overtime paid to cover the hole, so retention becomes a spending decision instead of a feeling.

Every restaurant operator knows turnover is high. It is the one operating condition of the business that gets stated as a given, usually with a shrug, usually followed by a comment about the labor market. What almost no operator can do is finish the sentence with a number. Turnover costs us... what? Per person? Per position? Compared to what?

The reason is not indifference. It is that a departure produces no bill. Nobody invoices you for a resignation. The costs it creates are all made of things you were already paying for: manager salary, trainer wages, overtime hours that look like a busy week, and the gap between what a new hire produces and what the person they replaced produced. Every one of those is inside a payroll number you already accepted.

The four costs hiding inside payroll

When a line cook quits, four things happen, and all four cost money.

Recruiting. Posting the role, screening whoever responds, the general manager's hours conducting interviews and no-showing candidates who never arrive. This is manager salary redirected from running the restaurant, and it is invisible because the manager is paid the same either way.

Onboarding. Paperwork, orientation, food handler certification, uniform, POS setup, adding the person to the schedule and payroll. An hour or two of an administrator's time, plus whatever the certification costs.

The productivity gap. This is the largest one and the least tracked. A new line cook does not produce what an experienced one produces on day one. Tickets run slower, waste goes up, plates come back, and a senior cook loses part of their own output to watching and correcting. That gap closes over weeks, not days, and every hour of it is paid at the full rate.

Cover overtime. Between the day someone leaves and the day their replacement is genuinely useful, the shift still has to be staffed. Somebody picks it up, usually at a premium, usually one of the good employees who is already carrying a lot. This is the cost with the nastiest second effect: covering for turnover is one of the reliable ways to create more of it.

Recruiting is buried in salary. Onboarding is buried in admin. The productivity gap has no line at all. Cover overtime does have a line, but it sits in a labor percentage where it reads as volume, not as a hole being plugged.

Why the labor percentage hides all of it

Most restaurant groups manage labor as a percent of sales, by location, by week. It is a good operating discipline and it is completely blind to this question. A location running 30 percent labor while backfilling three positions and a location running 30 percent labor with a stable crew look identical on the report. They are not remotely the same business.

Worse, the labor percentage can make the expensive situation look fine. A short-staffed location with people on overtime often runs an acceptable percentage, because sales are being covered by fewer, harder-worked people. The number says "in line." The reality is a crew burning down.

So the group compares its five locations on sales, on food cost, and on labor percentage, and never on the thing that is quietly separating them: what it costs each of them to keep a position filled.

A look at a restaurant group

Consider a five-location restaurant group doing about $10 million a year, casual full service, roughly 180 hourly employees across the group, an owner who is still in the restaurants several days a week, and a small back office running payroll and accounting on QuickBooks and a scheduling app.

Price out a single line cook departure.

Recruiting: posting and screening plus about four hours of a general manager's time at a loaded $32 an hour is roughly $160.

Onboarding: two hours of administrative time, plus certification and uniform, call it $120.

The trainer: a senior cook shadowing and correcting for roughly 40 hours across the first three weeks, at a loaded $24 an hour, and only partly productive during it. Count half of it as lost: $480.

The productivity gap: over the first six weeks the new cook averages perhaps 60 percent of the output of the person who left. On about 240 hours at a loaded $22, a 40 percent gap is roughly $2,100.

Cover overtime: five weeks of a hole in the schedule, about twelve overtime hours a week at an $11 premium over base, is roughly $660. This is conservative if the location is already thin.

Total: about $3,520 for one line cook, and a front-of-house position would land lower while a kitchen manager would land several times higher.

Now the group scale. If hourly turnover across 180 positions runs at roughly 80 percent annually, which many operators would recognize in their own numbers, that is about 145 separations a year. At a blended $3,800 across the mix of positions, the group is spending on the order of $550,000 a year refilling roles. Against $10M in revenue, that is more than five points, and none of it appears anywhere as a line called turnover.

What you would expect to see once it is measured by location is that the five sites are not the same. It would be unsurprising to find one location generating a disproportionate share of the departures, and to find that most of the separations cluster in the first 60 days, which points at hiring and onboarding rather than at pay. Those are different problems with different fixes, and neither is visible while turnover is one number for the group, or no number at all.

How the answer becomes available

The mechanism here is unglamorous: the events that make up a departure need to be connected to each other and to the position, rather than scattered across a scheduling app, a payroll file, and a manager's memory.

Hire date, separation date, position, location, and reason are almost always captured somewhere already. What is missing is that they are not joined to the overtime hours logged that month, to the manager hours spent hiring, or to the training period on the schedule. Once a position has a record and those things attach to it, cost per departure and time to full productivity stop being estimates and become figures the group actually holds.

Then the recurring work comes off people. Nobody assembles a turnover report. Tenure, first-60-day separations, overtime concentrated in one location, and repeat backfills of the same position surface as they happen. A manager who has now hired for the same slot three times in a year gets a flag rather than a reputation for bad luck.

The payoff is not the report. It is that a retention decision, a dollar an hour, a shift differential, a real training program, a schedule posted two weeks out, can finally be compared against a known number instead of argued about. A dollar an hour on a full-time line cook is roughly $2,000 a year. Against a $3,500 replacement cost, that math is arguable. Against a shrug, it is not arguable at all, which is why it usually loses.

How to start

You can build the first version of this from records you already keep.

1. Price one position properly. Pick your highest-churn role. Count recruiting hours, onboarding hours, trainer hours, the ramp period, and the overtime that covered the gap.
2. Count last year's separations by location and position. Payroll has the hire and termination dates. That is enough.
3. Multiply, and look at the concentration. The total matters less than which location and which role produce most of it.
4. Split first-60-day departures from the rest. Early exits are a hiring and onboarding problem. Later exits are usually a pay, schedule, or manager problem.
5. Let the tracking run itself. Set tenure, early separations, and cover overtime to surface as they occur, so the pattern shows up in a month rather than at year end.

The takeaway

A restaurant group at this size treats turnover as weather, something to endure rather than something to buy or not buy. It is not weather. It is a purchase of several hundred thousand dollars a year, assembled out of manager hours, trainer wages, slow tickets, and overtime, and it stays invisible because every one of those pieces is hiding inside a payroll number you already approved. Price one departure honestly, count how many you had, and find out which location is producing them. The number is likely to be larger than the retention spending you have been declining, and that comparison is the whole point.