Restaurant Staff Scheduling: How to Cover Every Shift and Every Run

Restaurant staff scheduling is the one weekly task where your labor cost and your service quality are decided at the same time, by the same decisions. Over-staff Tuesday and you’ve spent margin you can’t recover. Under-staff Friday at 7 p.m. and you’ve spent something harder to get back, which is the table’s willingness to return.

What makes it harder than scheduling most businesses is that a restaurant doesn’t have busy days, it has busy hours. The same Thursday can be dead at 4, slammed at 7, and dead again at 9:30. Staffing it as one block guarantees you’re wrong twice.

And most kitchens now carry a third stream of work on top of dine-in: takeout and delivery orders that arrive on their own rhythm and have to leave the building hot. This guide covers the whole week: forecasting covers by daypart, splitting front and back of house, handling the delivery runs, the labor cost percentage to aim at, and the rules you have to respect. If you want the general mechanics behind it first, the employee scheduling process lays out the six steps in a form that applies to any business.

The Bottom Line

• Schedule by daypart, not by day. Forecast covers in two- or three-hour blocks and stagger start times against them; a restaurant’s demand curve is far too steep for one shift per person per day.
• Full-service restaurant labor has been running at a median near 36.5% of sales, with more profitable operators holding closer to 34.2%, while limited-service and quick-service models sit nearer 25-31% (Restaurant Inventory Tools, 2026).
• Treat takeout and delivery as a third station with its own hours. Expecting the expo or a server to absorb it is how dine-in service slips on your busiest nights.
• Non-daytime work is the norm in this industry: 36.8% of leisure and hospitality workers usually worked a non-daytime schedule, more than double the 16.4% across all workers (U.S. Bureau of Labor Statistics, 2017-18).
• Publish two weeks out and hold the first week firm. Large fast-food and retail employers in several cities are legally required to do roughly this; for everyone else it’s the cheapest retention tool available.

What makes restaurant staff scheduling different

Three things, and each one breaks a scheduling habit that works elsewhere.

Demand moves by the hour. A 60-cover Friday night isn’t 60 covers spread evenly. It’s 12 at six o’clock, 34 between seven and half past eight, and the rest trailing off. Your schedule has to have its own peak, which means staggered starts rather than a single shift change.

Two halves of the house fail differently. Short a server and service slows down. Short a line cook and the kitchen backs up, which slows every table at once. They’re not interchangeable and they don’t scale at the same ratio.

Takeout and delivery run on a separate clock. Orders land when customers want them, not when the dining room is quiet. A 7:15 p.m. delivery order competes with the busiest dine-in moment of the week for the same pair of hands.

That third one is newer than the systems most restaurants use to schedule, which is why it’s the stream most often left unstaffed.

Build the restaurant schedule from the rush, not from opening hours

Start with the forecast and work outward. Opening time is the least informative number on the schedule.

Forecast covers by daypart

Break each day into blocks (open to 11, 11 to 2, 2 to 5, 5 to 8, 8 to close) and write down expected covers or sales for each. Your POS has this history; two months of it is plenty. If you don’t have the data, your own pattern knowledge is a legitimate input for a first pass.

Then convert covers into required bodies per station. A rule like “one server per 14 covers at peak, one per 20 off-peak” is crude but repeatable, and repeatable beats precise because you can correct it week to week.

Stagger the start times

Once you can see the curve, the schedule writes itself differently. Instead of four servers from 4 p.m. to close, you get two at 4, two more at 6, one leaving at 9. Same coverage at the peak, several hours of payroll saved at the edges.

Staggering is also where shift length gets decided. If your day has two peaks and a dead middle, two shorter shifts often beat one long one. And if you’re running early prep, late service and weekend extremes, consider whether a rotation serves you better than a fixed rota. Rotating shifts trade fairness gains against sleep costs, and the fairness argument is usually what decides it.

Assign the hardest shifts first

Friday and Saturday peak, the opening prep shift, Sunday brunch: fill those before anything else. The Wednesday lunch slot will find someone. The 6 a.m. prep shift will not, if you leave it until last.

Staffing front of house and back of house without over-staffing either

Build them as two linked schedules, not one list of names.

Back of house drives the ceiling. The kitchen determines how many covers you can actually produce in an hour. Staff it to the peak you expect to serve, then check that the dining room can’t seat more than the line can cook.

Front of house drives the experience. Servers, host, bar and bussers scale with the number of tables turning, and the host stand is the most commonly under-scheduled position in restaurants: one person who can control the pace of seating protects both halves of the house from the same bad hour.

Build in one flex body on peak shifts. Not an extra server, a person with a defined second job: expo who can bus, a bartender who can run food. Flexibility is cheaper than redundancy.

Where the two halves touch, at the expo window, is where delivery orders usually get dropped, which brings us to the stream most schedules ignore.

Who takes the delivery and takeout runs?

Answer it explicitly on the schedule, because the default answer is “whoever is closest”, and that’s your expo at 7:15 on a Friday.

Three arrangements work, and the choice depends on volume:

• Low volume (under roughly 15 orders a shift): add it to one named role per shift, with the hours to match. Someone owns packing and handoff. “The team handles it” is not an assignment.
• Steady volume: schedule a dedicated takeout station through the peak only. Two to four hours, not a full shift, costed as its own line in the forecast.
• High volume or your own drivers: it becomes a separate stream with its own staffing, including the people who pack and the people who drive.

If you run your own delivery, add the step restaurants most often forget: packing and checking time before the driver leaves. A 20-minute window between the last plate and the van pulling out is work, and it needs hours on the schedule the same way prep does. Driver hours also scale with distance rather than order count, so a night with the same order volume spread wider costs more delivery labor.

Where staff are driving their own routes, hours get awkward to predict, and a degree of self-scheduling inside clear rules absorbs it better than a fixed rota does. That’s the mechanism behind flexible work hours for hourly teams, which is less a policy than a swap system with guardrails.

Restaurant labor cost percentage: what to aim for

Labor cost percentage is total labor cost divided by sales for the same period. Loaded labor, not just wages: payroll taxes, benefits and employer-paid insurance all count, and together they add meaningfully to the headline wage.

Benchmarks vary hugely by concept, so the only number worth comparing yourself to is your own segment’s:

Source: Restaurant Inventory Tools, 2026 benchmarks by concept.

Two habits make the number useful rather than decorative. Check it weekly, because a monthly figure describes a problem you can no longer fix. And price the schedule before you publish it. The cost of next week is fully knowable on Thursday, and far cheaper to change then than to discover on payday.

Watch the overtime creep too. Someone scheduled 38 hours whose shift reliably runs 20 minutes long is on overtime every week, and nobody notices until it’s paid. Most scheduling tools flag this while you’re still editing, which is one of the stronger arguments for using one, covered in detail in how to choose shift scheduling software.

Handling shift swaps, call-outs and the Saturday problem

Changes are not a failure of the schedule. They’re the normal operating condition of a restaurant, and the question is only whether you’ve designed for them.

• Let staff swap within rules. Same role, same qualification, submitted before a cutoff, approved by a manager. This removes most of the manager hours that scheduling actually costs.
• Keep an open-shift list. People who want extra hours will claim them if claiming is easy. That’s your first call on a Saturday morning, not a phone tree.
• Rotate the shifts nobody wants. If Friday close always lands on the same two people, you have a fairness problem that becomes a hiring problem. Rotation is the standard fix.
• Record every change in one place. The schedule that lives in three group chats is not a schedule, and the version confusion costs you a no-show eventually.

Scheduling rules restaurants have to follow

This is the area where restaurants carry more obligation than most small businesses, and it varies by jurisdiction. Check your state and city. The list below is what to check for, not legal advice.

Fair workweek and predictive scheduling rules. Several cities and states require advance notice of schedules plus premium pay for late changes, and restaurants are usually the target. Oregon’s statewide law covers retail, hospitality and food service employers with 500 or more employees worldwide and requires the written schedule at least 14 calendar days in advance (Oregon BOLI, retrieved 2026-10-02). New York City’s fast-food rules also run on 14 days, with retail at 72 hours. Most independents fall under the headcount thresholds, but the standard those laws set is a good one to meet voluntarily.

Clopening restrictions. Where these rules exist, they typically require 10 or 11 hours between the end of one shift and the start of the next, with premium pay if an employee consents to less. Scheduling the same person to close at 11 and open at 7 is bad practice regardless of whether it’s prohibited where you are.

Minors. Hour and late-night restrictions for employees under 18 are state law everywhere and are easy to breach accidentally when you’re covering a call-out at 9 p.m.

Breaks and rest periods. Required length and timing differ by state. If your schedule assumes a cook works a six-hour push without a break, check whether that’s legal before it becomes a claim.

The restaurant scheduling mistakes that cost the most

• Scheduling by headcount habit. “Four servers on Saturday because we always have four” isn’t a forecast, it’s a tradition with a price tag.
• One shift per person per day. The demand curve has a peak. A schedule without one is paying for the trough.
• Leaving takeout and delivery unassigned. The cost lands on dine-in service during your highest-revenue hour.
• Under-staffing the host stand. The cheapest position to cut and the one that controls how badly the rest of the night goes.
• Forgetting prep and pack time. Hours that produce no covers directly still have to exist on the schedule.
• Publishing late, every week. Short notice raises call-outs and turnover, and in this industry turnover is the most expensive line you don’t see on a P&L.

Frequently asked questions

How do you schedule restaurant staff efficiently?

Forecast covers by daypart, convert those into bodies needed per station, stagger start times to match the peak, fill the hardest shifts first, price the week before publishing, and publish at least two weeks out with a clear process for swaps and call-outs.

What is a good labor cost percentage for a restaurant?

It depends on the concept. Full-service operations have been running near a 36.5% median with more profitable ones closer to 34.2%, while quick-service and fast-casual models typically sit between 25% and 31% (Restaurant Inventory Tools, 2026). Compare yourself to your own segment rather than to a single industry figure.

How far in advance should a restaurant post the schedule?

Two weeks. Large fast-food and retail employers in several jurisdictions are legally required to give roughly that much notice, and for independent restaurants it reduces call-outs and turnover at no cash cost.

Who should handle delivery and takeout orders during a rush?

Someone named on the schedule, with hours allocated for it. Below about 15 orders a shift, attach it to one existing role; above that, schedule a dedicated takeout station for the peak hours only. Leaving it to whoever is nearest pulls your expo away from dine-in at the worst moment.

Should restaurant staff schedules be the same every week?

A stable base helps retention, because people can plan their lives around it. Keep core shifts consistent for reliable staff and flex the edges, adding hours at the peak and trimming the trough, rather than rebuilding the whole week from scratch each time.