Restaurant Employee Turnover Rate: What It Means
Where the industry turnover figure comes from, why benchmarking against it misleads, how to calculate your own, and the two splits that show what to fix.
Every article on this topic opens with the same number. Restaurant turnover tops 75 percent. Quick service runs higher still. Workers quit at about twice the national rate.
The numbers are roughly right and close to useless for running a restaurant. A percentage that combines quits and firings, counts seasonal endings alongside real losses, and describes an industry rather than a building cannot tell you what to change on Monday. Two restaurants with identical staffing can report 60 percent and 140 percent depending only on how each one counts.
This guide covers where the industry figure comes from, what federal data does and does not say, and how to calculate your own without fooling yourself. Then the two splits that turn a percentage into a decision: separating quits from discharges, and splitting departures at the 90-day mark.
After that come the parts nobody writes for a single-unit operator: what turnover really costs at your size, why tipped pay and scheduling are the same problem, and where to start. I build the employee records and first-week onboarding plans this rests on at FirstHR. This is general information rather than legal advice.
- Sign the offer letter
- Sign the employee handbook
- Workplace Safety course
What a Turnover Rate Actually Measures
Turnover rate is the number of employees who left over a period, expressed as a percentage of average headcount over that same period. It measures departures against staffing size, not the share of your people who walked out.
That distinction between departures and individuals is the one that changes how you read every figure on this page, and it is covered in detail further down.
Turnover is also easy to confuse with a related term. Turnover and attrition are used interchangeably in most restaurant content, but attrition more properly describes positions that go unfilled by choice while turnover describes positions you refill. In a restaurant almost everything is turnover, since you replace the person.
Where the 75 Percent Figure Comes From
The honest answer is that it is a federal number, though not a restaurant number. According to the Bureau of Labor Statistics Job Openings and Labor Turnover Survey (2018 annual rates), total separations in accommodation and food services equaled 74.9 percent of average employment that year, which rounds to the familiar 75 percent.
The National Restaurant Association reported that rate as turnover in the restaurants-and-accommodations sector. The part that matters sits in the association's footnote: the figure covers hotels too, because the bureau does not report restaurants alone. It also counts every separation, from quits to firings to the end of a seasonal job.
That does not make it useless. As an order of magnitude it holds up: restaurant turnover runs far above the economy-wide norm. Private industry data, which can see inside the sector, adds that quick service runs higher than full service and that management is substantially more stable than hourly staff.
What it cannot do is serve as a benchmark. When you compare your 90 percent to the industry 75 percent, you are comparing a number you calculated one way to a number somebody else calculated another way, from a sample that probably looks nothing like your restaurant. The gap between the two figures tells you almost nothing about your operation.
The latest reading of the same measure gives the current neighborhood. According to the Bureau of Labor Statistics Job Openings and Labor Turnover Survey (2025 annual averages), total separations in accommodation and food services ran at 5.5 percent of employment a month, or roughly 66 percent of average employment across the year.
Private sources land somewhere else again, which is the practical problem with any benchmark. Market data in the National Restaurant Association's 2026 hiring and staffing report put restaurant and foodservice turnover at 122 percent for 2025, almost double that federal reading for the same year.
What Federal Data Actually Says
The federal data behind that number is the Bureau of Labor Statistics Job Openings and Labor Turnover Survey, which is both more reliable and less specific than private industry figures.
The survey reports hires and separations by industry, including Accommodation and Food Services. It does not publish a restaurant turnover rate as such, and it does not break out restaurants from hotels.
What the survey does publish is monthly rates for quits, for layoffs and discharges, and for total separations, which is every departure from the payroll in the reference month and the closest published equivalent to a turnover count. Two things in that data stand out for anyone running a restaurant.
The first is that accommodation and food services posts the highest quits rate of any industry tracked. According to the Bureau of Labor Statistics Job Openings and Labor Turnover Survey (2025 annual averages), the sector quit at 4.2 percent of employment a month, against 2.2 percent across the private sector and 2.6 percent in retail trade, the next highest.
The second is more useful. Layoffs and discharges in the sector ran at 1.1 percent of employment a month in the same data, matching the all-industry figure and sitting below the 1.2 percent private-sector rate. The sector does not let people go at an unusual rate.
Put those together and the sector-level picture is specific. The industry does not have a firing problem. It has a leaving problem. Roughly three in four departures in accommodation and food services are people choosing to go, which is why an operator whose own numbers look different has learned something important about their own restaurant rather than about the industry.
Quits Are Not Firings, and Mixing Them Hides the Problem
Quits and firings point at different problems, and a turnover figure that mixes them hides both. This is the first of the two splits, and the one that costs least to do and returns most.
The reason this matters at a single restaurant is that the two categories point at opposite ends of the process. A restaurant with high quits has a retention problem: people arrive, work, and decide to leave. A restaurant with high discharges has a hiring problem: you keep selecting people who do not work out, or you are hiring for a role that is set up to fail.
Those need completely different responses, and a combined turnover figure makes them indistinguishable. An operator who discovers that 60 percent of their departures are discharges should stop reading about retention strategies entirely and go look at how they interview, what they tell candidates the job is, and what happens in week one.
One practical note. No-shows and abandonment sit awkwardly between the two, since the employee effectively quit and the paperwork records a termination. Pick one treatment, write it down, and apply it consistently, because in a restaurant this category is large enough to swing the split on its own.
How to Calculate Your Own Rate
Divide departures by average headcount and multiply by 100. The formula is simple, and the decisions around it are where accuracy lives.
Nick Anisimov, FirstHR founder
Why Benchmarking Against the Industry Misleads
Benchmarking against the industry misleads because of six methodological choices, all defensible and all made differently by different sources. Several of them can move a reported rate by double-digit percentage points.
| Choice | One way | The other way | Effect on the number |
|---|---|---|---|
| Seasonal staff | Included, endings counted as departures | Excluded entirely | Can swing a seasonal operation by 30 points or more |
| Counting basis | Bodies on the payroll | Full-time equivalent | Many short shifts inflate the body count and lower the rate |
| Discharges | Counted with quits | Quits only | About a fifth of sector departures in 2025 federal data |
| Period | Calendar year | Rolling twelve months | Changes which season sits at the boundary |
| Rehires | Counted as a new hire and a new departure | Netted out | Meaningful where staff return between seasons |
| Trial shifts and no-shows | Counted as employment then departure | Not counted as ever employed | Can add double digits at a high-volume operation |
Headline figures rarely travel with these choices attached. The federal figure counts every separation, seasonal endings included, against everyone on the payroll, while private surveys make their own calls. That is the whole argument against benchmarking: you are comparing your figure to a figure built differently, and the difference between them is more likely to reflect method than performance.
What is worth comparing is your restaurant to itself. Same formula, same treatment of seasonal and discharges, measured monthly. A rate that moved from 88 to 71 on a consistent method is a real result. A rate of 71 against an industry 75 is a coincidence.
Turnover Concentrates in Positions, Not People
Turnover usually sits in a few scheduled positions rather than spreading across your people. This is the most useful idea on the page, and it is almost absent from published content on the topic.
Concentration is the normal pattern rather than the exception, and once you see it the diagnostic question changes completely. It stops being why is our turnover high and becomes why does this particular slot keep emptying.
That question about one slot usually has a concrete answer: the shift pattern attached to it, the station, the person who supervises it, or the fact that it is the role everyone is promoted out of.
Counting it is straightforward. For each scheduled position, count how many different people occupied it over the year. Two people through a dish slot is normal. Five is a signal. The table takes twenty minutes to build from your schedule and it will point at one or two positions carrying a disproportionate share of the total.
The reason this matters practically is cost. Fifteen departures spread across fifteen positions is a diffuse problem needing broad action. Fifteen departures concentrated in four positions is four problems, and fixing the worst one might remove a third of your turnover without touching pay for anyone else.
Split Every Departure at 90 Days
Split departures into those inside 90 days and those after, because the two have different causes and different fixes. This is the second split, and the one that separates the fixable from the expensive.
| Left within 90 days | Left after 90 days | |
|---|---|---|
| What it usually means | The job was not what they expected, or nobody trained them | Pay, schedule, progression, or the manager |
| Who controls it | You, almost entirely | You, but at higher cost |
| Typical cause | The hiring conversation or the first week | Accumulated experience of working there |
| Cost to fix | Low, often just process | Real money or real management change |
| Where to look | The interview, the first four shifts, who trained them | Scheduling, pay, the supervisor, the path upward |
| Speed of feedback | Weeks | Months to a year |
Early departures are the cheapest turnover you will ever fix, and they are usually the largest single block in a restaurant. Somebody who leaves in three weeks did not develop a grievance about career progression. They discovered the job was different from the description, or worked four shifts without anyone showing them the system, or were scheduled for hours they had already said they could not do.
All three of those are process failures inside your control, and none of them costs money to fix. Being accurate in the interview about the hard parts, running an actual restaurant onboarding checklist for the first week, and confirming availability before the first schedule goes out will remove a meaningful share of early departures at a restaurant that currently does none of them.
Departures after 90 days are the harder category and worth reading differently. Somebody who stayed nine months and left has told you something about sustained conditions rather than about first impressions. That is where pay, scheduling, and the supervisor relationship actually live, and where the first 90 days framing stops being the answer.
If You Run Under Thirty People
A caution that applies to most independent restaurants and appears nowhere in industry content, because industry content is written for operators with hundreds of staff.
At small headcounts the percentage is statistically noisy to the point of being misleading. With twelve employees, one departure is more than eight percentage points. Two people leaving in the same month for unrelated personal reasons produces a rate that looks alarming and means nothing. Chase it and you will make changes in response to noise.
The compensating advantage at this size is substantial and worth naming. You know every person. You can find out why somebody left by asking, and you will get a more honest answer than any survey instrument produces at a large operation, provided you ask after they have gone rather than during the notice conversation.
A short, genuinely low-stakes exit conversation is more informative at fifteen employees than analytics is at fifteen hundred.
What Turnover Actually Costs at Your Restaurant
Published replacement costs are quoted constantly and are worth treating with care, because they come from particular studies of particular operations and are frequently reproduced without their original date.
The components are more useful than any headline figure, because you can price them for your own restaurant in an afternoon.
Two observations that change the arithmetic for an independent operator.
The largest cost is almost never recruiting. It is the training period, when you pay two people to produce what one of them would produce alone, and the coverage gap before that. Operators who price turnover from job posting fees are measuring the smallest line.
The research points the same way. According to a Cornell University Center for Hospitality Research study of 33 properties (2006), lost productivity from inexperienced new hires was the largest of the five cost categories measured, reaching up to 70 percent of the total. That study covered hotels rather than restaurants, but the shape of the cost carries over.
And the cost is not linear with the number of departures, because concentrated turnover compounds. A position that turns over five times means five training periods, five coverage gaps, and a stretch of the schedule that is permanently unsettled, which is itself a reason the people around it start looking. That is the mechanism by which one bad slot becomes two.
Why Restaurant Turnover Runs So Far Above Everything Else
The structural reasons are worth understanding because they tell you which parts you can change and which you cannot.
| Driver | Why it operates in restaurants | Can a single-unit operator change it? |
|---|---|---|
| Pay relative to physical demand | Hard physical work at or near the wage floor for many roles | Partly, and it is the most expensive change to make |
| Income that varies by shift | Tipped earnings depend on which shifts you get, not just how many | Yes, through how sections and shifts are allocated |
| Unpredictable scheduling | Demand varies and schedules are often posted late | Yes, and this is usually the cheapest real improvement |
| Low switching cost for the worker | A server can change employers without changing career, skill, or commute | No, but it raises the value of everything else |
| Young, part-time workforce | Many staff treat the role as temporary from the outset | No, though it changes what retention means |
| The direct supervisor | One shift lead can account for a whole station of departures | Yes, and it is frequently the actual cause |
| No visible path upward | Small operations have few promotions to offer | Partly, through scope and skills rather than titles |
The fourth row is the structural one people underrate. In most industries leaving a job means disruption to a career. In a restaurant it means a different building. That does not mean retention is hopeless; it means the ordinary reasons people tolerate a job they dislike do not apply here, so the things you can control matter proportionally more.
The sixth row is the one most likely to be your actual answer and the least likely to appear in your data. If departures cluster around one station or one shift lead, that is a management issue wearing an industry costume, and the position table above will surface it.
The Tipped Pay Problem Nobody Frames Correctly
For tipped roles there is a structural feature that most retention content skips: a server's income is determined less by their hourly rate than by which shifts and sections they are given.
Two servers on identical wages can earn substantially different amounts in the same week because one worked Friday and Saturday evenings and the other worked Monday and Tuesday lunches. That is not a pay policy decision anybody wrote down. It is the schedule, made by a manager, usually without anyone treating it as a compensation decision.
The legal layer compounds it. Under federal rules an employer may take a tip credit against its minimum wage obligation only under specific conditions, and time spent in a genuinely separate non-tipped occupation is treated differently. The Department of Labor sets out that framework, including tip credit conditions, tip pooling, and dual jobs, in its tipped employees regulations.
One change to that framework matters for how you schedule. The 2021 rule that barred a tip credit for directly supporting work beyond 20 percent of the workweek or 30 continuous minutes was vacated in court.
The Department of Labor removed those limits from the dual jobs regulation in an amendment dated December 17, 2024. No percentage limit on supporting work survives in the federal text.
State law can remove the question entirely. Seven states require the full state minimum wage before tips, with no credit at all: Alaska, California, Minnesota, Montana, Nevada, Oregon, and Washington. The Department of Labor state table effective July 1, 2026 lists Guam alongside them.
Where this touches turnover directly: a server scheduled heavily on prep and cleaning is earning less per hour than one scheduled on the floor, and if that work amounts to a separate occupation it raises a dual jobs question at the same time.
Scheduling Predictability Costs Almost Nothing
If pay is the most expensive option and management change is the slowest, scheduling is the one that costs almost nothing and moves the number fastest.
The mechanism is straightforward. An hourly worker cannot plan childcare, a second job, a class, or a rent payment against a schedule they receive three days out and that changes after posting. That is a bigger source of pressure than an extra fifty cents an hour, and it is why people leave restaurants for jobs that pay the same.
The research backs that mechanism directly. According to The Shift Project at Harvard Kennedy School (ILR Review, 2021), a panel of 1,827 hourly retail and food service workers showed schedule instability to be a strong, robust predictor of turnover for the roughly one third of workers with relatively unstable schedules.
In that study, job satisfaction explained slightly less than half of the link between schedule instability and turnover, and work-family conflict explained another quarter.
The last step is a genuine compliance point rather than a good practice. Coverage varies by jurisdiction and by employer size, and our guide to predictive scheduling laws covers where they apply.
What makes scheduling work at a restaurant is consistency rather than sophistication: the same notice and the same rules, every week.
What to Do First
In order, cheapest and fastest first. Most restaurants attempt these in reverse.
The first two steps are record-keeping rather than decisions, and the workbook below is the shape they take.
| A | B | C | D | E | F | G | H | I | |
|---|---|---|---|---|---|---|---|---|---|
| 1 | Name | Position | Hire date | Last day | Days employed | Left within 90 days | Quit or discharged | Reason given | Reason we believe |
| 2 | Example: A. Duran | Dishwasher | 2026-03-02 | 2026-03-24 | 22 | Yes | Quit | Found closer job | Scheduled only closing shifts |
| 3 | Example: R. Kim | Server | 2025-04-14 | 2026-05-30 | 411 | No | Quit | Moving | Lost weekend sections |
| 4 | |||||||||
| 5 | |||||||||
| 6 |
Three sheets for the three things worth knowing. The departures log captures the fields that make a departure interpretable rather than just counted, including the two columns most restaurants omit: days employed and what you actually believe the reason was, which is frequently not the reason given.
The position sheet is where concentration shows up. The rollup produces the percentage, calculated consistently, with quits and early departures tracked alongside it so the headline number never travels alone.
What Holds People Past the First Year
Most of the list above targets the first 90 days. Keeping a two-year server takes something different: something to learn, something to be trusted with, and a reason to believe next year looks better than this one. Restaurants are unusually bad at this because the only visible promotion is into management, and a single-unit operation has one of those to give.
Cross-training is the tool most independent restaurants already own and do not use. A server who can also run expo, or a line cook who can handle two more stations, becomes harder to replace, easier to schedule around, and eligible for the shifts that pay. Attach a small differential to each station somebody signs off on and the path is real without inventing a title.
Recognition matters less as a program than as specificity. Naming what somebody did well, in front of the people whose opinion they care about, costs nothing and gets remembered. What does not work is recognition layered over an unfixed schedule, because it reads as a substitute for the thing people actually asked for.
Where Operators Get This Wrong
Six mistakes repeat across restaurants of every size, and a few quieter ones follow them.
Benchmarking against the industry figure is first. It was counted by someone else's rules, so a gap between theirs and yours says more about method than about how you run the place.
Reporting one combined number is second. High quits signal a retention problem and high discharges a hiring problem, and one blended figure hides which of the two you have.
Ignoring the 90-day split is third, and it is the most expensive omission. Early departures are the largest block and the cheapest to fix, and they are invisible in an annual percentage.
Treating turnover as spread evenly across staff is fourth. It concentrates in positions, and the position table finds in twenty minutes what a retention program spends months not finding.
Chasing the monthly percentage at a small restaurant is fifth. Under thirty staff, that number is mostly noise, and reacting to noise produces changes that were never needed.
Reaching for pay first is sixth. It is the most expensive option, it is often not the binding constraint, and it will not hold if the first week and the schedule are still broken.
Then the quieter ones. Changing the calculation method mid-year, so the trend becomes uninterpretable. Never asking departing staff anything, at the size where asking works best. Treating schedule allocation as an operational task rather than the pay decision it is in a tipped house. And running exit conversations during the notice period, when nobody has any reason to be candid.
None of this needs an HR function. It needs a departure log with three extra fields, a position count once a year, and a first week that somebody owns. That is the same infrastructure that makes the rest of HR at a small business work without a specialist.
The restaurants that fix this are usually the ones that started measuring differently rather than spending more.
Frequently Asked Questions
What is the average turnover rate for restaurants?
The figure repeated most widely across the industry is around 75 percent a year, with quick service cited higher still. That number is the federal rate of total separations for the whole accommodation and food services sector in 2018, 74.9 percent, so it includes hotels, because the Bureau of Labor Statistics publishes no rate for restaurants alone. Its monthly rate across 2025 averaged 5.5 percent of employment, which works out to roughly 66 percent over the year. A 2026 hiring report from the National Restaurant Association, drawing on market data, gives 122 percent for restaurant and foodservice turnover in 2025. The spread is the point. Treat 75 percent as an order of magnitude and a poor benchmark, because two restaurants with identical staffing realities can report very different percentages depending on how each one counts.
How do you calculate restaurant turnover rate?
Take the number of people who left during the period, divide it by your average headcount over that same period, and multiply by 100. Average headcount is typically the headcount on the first day plus the headcount on the last day, halved. For a yearly rate, add up every departure across the twelve months and use the average staffing level for the year as the divisor. Three decisions change the answer substantially and should be made once and kept consistent: whether seasonal staff are included, whether you count people or scheduled positions, and whether discharges are counted alongside voluntary quits.
Why is restaurant turnover so high?
Several structural reasons compound. Pay is low relative to physical demand and, for tipped roles, varies by shift rather than being predictable. Schedules change week to week, which means income changes week to week. The work is physically and emotionally hard, often at unsociable hours. The workforce skews young and part-time, with many people treating the job as temporary from the start. And the alternatives are close by: a server can change employers without changing career, commute, or skill set, which makes leaving unusually easy compared with most industries.
What is a good turnover rate for a restaurant?
There is no single number, and chasing the industry average is the wrong target. A more useful standard has three parts. Turnover within the first 90 days should be low, because early departures are usually the result of a hiring or onboarding failure you control. Turnover should not be concentrated in one or two positions, because concentration signals a specific fixable problem rather than industry conditions. And voluntary quits should outnumber discharges by a wide margin, since a high discharge count is a hiring problem wearing a retention costume.
How much does it cost to replace a restaurant employee?
There is no dependable published number for an independent restaurant. The figures in circulation come from individual studies of particular operations, several of them more than a decade old, and they are usually repeated without their date. A better exercise is to price your own components: time spent posting and screening, the hours a manager spends on interviews and paperwork, training shifts where both the trainee and the trainer work below full output, coverage while the position sits empty, and the mistakes and slower service that follow. Academic work on hospitality turnover points consistently at the output lost while new hires are still inexperienced as the biggest of those components, which is why an estimate built from job posting fees understates the real cost badly.
Does 75 percent turnover mean three quarters of my staff left?
No, and this is the most useful thing to understand about the number. Turnover measures departures against average headcount, not the share of individual people who left. A twenty-person restaurant with fifteen departures in a year usually has not lost fifteen different long-standing employees. It has more often had four unstable positions that turned over repeatedly while the people in the other sixteen slots stayed put. Same percentage, completely different problem, and only the second description tells you where to intervene.
Which restaurant roles have the highest turnover?
Hosts, counter staff, and servers turn over fastest, dishwashers and prep workers are not far behind, and management is far more stable. A 2026 hiring report from the National Restaurant Association, using market data, put 2025 turnover at about 293 percent for hosts and hostesses, 173 percent for fast food and counter workers, 131 percent for waiters and waitresses, 124 percent for dishwashers, and 104 percent for food preparation workers, against about 46 percent for food service managers. Industry survey data also shows quick service turning over faster than full service. Rather than relying on those industry patterns, count turns per position at your own restaurant: how many people passed through each scheduled slot over the year. That single table is more actionable than any published role benchmark, because the answer is frequently a specific shift pattern rather than the role itself.
How do I reduce turnover in my restaurant?
Start by finding out where it is concentrated rather than launching a general retention effort. Count turns per position, split departures at the 90-day mark, and separate quits from discharges. Early departures usually point at the hiring conversation or the first week, which are cheap to fix. Concentrated departures in one position usually point at a shift pattern, a station, or one supervisor. Only once those are addressed does broad action on pay and scheduling predictability earn its cost, and by then you will know which of the two matters more at your restaurant.
Should I count seasonal staff in my turnover rate?
Count them, but track them separately. Excluding seasonal departures flatters the number and hides real costs, since every seasonal hire still consumes training time and management attention. Including them without separating them makes your figure incomparable across years in which your season was longer or shorter. The practical approach is one figure including everyone for cost purposes, and a second excluding planned seasonal endings for retention purposes, with the method written down so the same rule applies next year.
Make HR the easy part of running your business
Onboarding, e-signatures, training and core HR in one place. Up and running in 5 minutes.
- 7-day free trial
- No credit card required
- Cancel anytime