In March 2026, the quits rate in accommodation and food services was 4.3%. For all private employers it was 2.2%. Hiring ran hot too: 6.2%, the highest of any private industry, per the Bureau of Labor Statistics.
Most operators price that churn as a hiring bill: the job post, the interviews, the paperwork. It is a real number. It is also incomplete. The restaurant employee turnover cost that is easy to miss is the training reset.
A note on the data
Those are monthly rates, and the category includes hotels as well as restaurants, so treat it as the nearest public read on the labor pool you hire from. By August the quits rate had eased to 3.5%, against 2.1% for all private employers. Lower. Not low.
01What the training reset actually is
A written standard is a floor. The real standard is what your people do on a Friday night when nobody is reading the manual.
That standard lives in people. Your best server knows how long to leave a table alone after the entrees land. Your lead line cook knows when a ticket is about to go sideways. Most of it was never written down, because it never had to be. It was passed hand to hand. That is what restaurant staff training really is: a standard moving from one person to the next.
So when someone quits, a version of your standard leaves with them. When their replacement starts, the new hire gets a version of it from whoever has ten minutes. Not the best version. The nearest one.
That is the training reset. Every quit resets part of your standard to whatever the next trainer remembers.
02What walks out the door with one server
Picture a server we will call Dana. She is a composite, not a real person. She has worked the floor at one of your restaurants for three years. She gives two weeks’ notice on a Tuesday.
On paper, you have one open position. In practice, here is what goes with her:
- How she reads a four-top in the first minute: who is hosting, who is in a hurry, who wants to be left alone.
- The order she sequences drinks, starters and entrees when the kitchen is slammed.
- What she tells new hires about the pass: who to ask, who not to bother, and when.
- Which regulars want which table, and how she handles each of them.
- The small habits that make a table feel looked after, like resetting before the next guest is seated and refilling without being asked.
The replacement starts next week. Her first three shifts are shadowing. Who she shadows depends on the schedule: a strong server on a good night, a tired one on a double, or a manager covering a hole. Each will teach what they do. None of them is wrong. They are just not the same.
Now run that across every position, every quit and every location.
03The restaurant employee turnover cost your budget leaves out
Hiring costs have invoices. The training reset does not, so it hides in four places:
The server who shadows a new hire is not fully working their own section.
Covering, correcting and answering the same questions again.
Slower tables until the new habits settle.
Which standard the new hire learned, and from whom.
The first three are hours you can count. The fourth shows up later, as a regular who says one location is not like the other.
This is why the usual turnover math undercounts. A hiring estimate stops the day the new hire clocks in. The reset starts that day and runs for weeks. And it repeats: every quit starts a new one.
04Run this self-audit with your own numbers
There is no industry benchmark in this exercise, on purpose. Your numbers are the benchmark. For each location, pull three things for the last 12 months:
- Who left. Every hourly front-of-house employee you replaced, by role and location. Payroll or HR exit data works.
- Training hours each replacement needed. Shadow shifts, trainer time and manager time. Use schedules and payroll. Estimate where you must and mark the estimates.
- Who delivered it. The name of each person who trained each replacement.
Then fill in one row per location:
| Location | Quits in last 12 months | Replacements trained | Avg training hours per replacement | Total training hours | Different trainers used | Of those, designated trainers |
|---|---|---|---|---|---|---|
| [Location 1] | ||||||
| [Location 2] | ||||||
| [Location 3] | ||||||
| All locations |
Read it three ways:
- Total training hours is the size of the reset. Multiply by what those hours cost you.
- Different trainers used is how many versions of your standard each location is teaching.
- Designated trainers shows how much of your training is left to whoever was free.
Compare locations to each other, and this year to last year. The gap between locations is the finding.
Do not wait for a perfect dataset. A rough audit of one location, with estimates marked as estimates, tells you more than no audit. If the trainer names are hard to find, that is a result too. It means nobody owns the reset.
05Why trainer-to-trainer variation becomes drift
Every trainer teaches the version they hold. Some hold it well. Some picked it up in a rush from someone who picked it up in a rush.
One reset is a small change. Dozens of resets are not. Each hire is trained by someone who was trained by someone else, and details drop out or get added along the way. It works like a photocopy of a photocopy.
Now spread that across locations. Location A trains new hires with its best server. Location B trains with whoever is available. Location C had two long-tenured servers who carried the standard, and both left in the same quarter. Same brand. Same manual. Three different restaurants.
The GM often cannot see it, because the GM was not in the room for most of the training. Drift does not announce itself. It shows up in guest comments, in uneven pacing between locations, and in the question of why one restaurant keeps outperforming its twin.
It follows that a location with more quits goes through more resets, and has more chances to drift. That is logic, not a statistic. Your audit table will show whether it holds in your group. More on how standards slip between locations: Restaurant Standards: What a Standard Is and How to Write One a New Hire Can Follow.
06Three things to do this month, no new software
You do not have to reduce restaurant turnover to start. You have to make each reset cleaner.
- Name your trainers. Pick one or two per location, tell them why, and pay them for it. Training by “whoever is free” stops being the default.
- Write the one-page standard with your best server. Sit with them for an hour. Ask what they do in the first minute, the first ten minutes and the last five of a table. Put it in their words, in plain sentences. Hand it out in every first week.
- Put your trainers in one room. Once a month, trainers from every location walk through the same table together and agree on how it should go. Forty-five minutes. The differences surface right there.
None of this needs a platform or a budget line. It needs a calendar invite and a conversation with your best people. Start with one location, then run the self-audit again in a quarter and see whether the number of different trainers has come down.
Pick the location where people leave most often. List the last five people you hired to replace someone. Next to each name, write who trained them. Count the different names. Five names means five versions of your standard.
Think of the last server who left your group. Who trained the person who replaced them? Tell us in the comments.
07Leave a comment
08FAQ
What counts as restaurant employee turnover cost?
It has two parts. The visible part is hiring: job posts, interviews and onboarding paperwork. The hidden part is the training reset: trainer hours, manager time, the new hire’s slower first weeks and the version of your standard the replacement learns. Most budgets capture the first part. The second takes a simple audit to see.
How do I calculate the training cost of replacing a restaurant employee?
List everyone who left in the last 12 months. For each replacement, write down the hours spent on shadow shifts, trainer time and manager time. Multiply those hours by what you pay for them, and note who delivered the training. Compare locations. The totals are yours, so no outside benchmark is needed.
Why does restaurant staff training differ between locations?
Training is passed hand to hand, so each trainer teaches the version they know. When one location uses its strongest server and another uses whoever is free, new hires learn different versions of the same standard. Over several rounds of turnover, those small differences add up to locations that run differently under one brand.
How can I reduce restaurant turnover?
Start by finding where people leave: by location, role and how long they stayed. Then look at how those employees were trained and who trained them. This guide does not promise a fix. It shows how to measure the training side of turnover, so you can decide where to look first.
What is the quits rate?
In the Bureau of Labor Statistics JOLTS survey, quits are generally voluntary separations initiated by the employee, excluding retirements and transfers. The quits rate is the number of quits during the month as a percent of employment. It is a monthly figure, so a 4.3% rate describes one month, not a full year.


