Labor is the line you can move fastest. Here’s how to schedule to your sales forecast, staff each shift to demand, and keep labor cost in range without leaving the floor short.
July 9, 2026
After food, labor is usually your biggest cost — and it’s the one you can move the fastest. You can’t change your rent this week, and re-negotiating food cost takes months. But the schedule you post on Thursday changes what you spend next week. That’s the good news and the trap: cut too hard and Saturday night falls apart, guests wait, servers burn out, and reviews suffer. Overstaff and you watch labor eat a margin you can’t get back. Good scheduling isn’t about being cheap — it’s about matching the people on the floor to the guests coming through the door, shift by shift. This guide walks through how experienced operators build a schedule that protects both the P&L and the service, and the mistakes that quietly cost the most.
Most owners think of the schedule as a logistics chore — who’s available, who wants Friday off, who called in. It’s that, but underneath it’s the single biggest weekly lever you have on profit. Labor typically runs somewhere in the range of 25 to 35 percent of sales for full-service restaurants, a bit lower for counter-service and quick-serve. A few points either way is the difference between a healthy month and a break-even one. The point isn’t to squeeze every hour out of the schedule; it’s to spend your labor where it actually earns — on the shifts and stations that carry the guest experience — and to stop bleeding it on slow mid-afternoons that don’t need three servers.
The most common scheduling mistake is copying last week and nudging it. That bakes in whatever was wrong last week and ignores what’s actually coming. Build the schedule off a sales forecast instead. Pull the same weekday from the last several weeks, layer in the seasonal trend, and adjust for anything you know: a long weekend, a local event, a patio day, a slow week after the holidays. Your POS already holds this history — hourly sales, covers, average check by daypart — so you’re not guessing. When you can see that last three Tuesdays did most of their business between 6 and 9, you stop scheduling a full team at 4:30 and start staffing the hours that pay for themselves.
A forecast tells you the shape of the day; the schedule turns it into people. Break the day into the parts that actually behave differently — open, lunch, the mid-afternoon lull, the dinner build, close — and staff each to its own curve rather than blanketing the whole day with one headcount. Stagger start and end times so people arrive as volume builds and cut as it fades, instead of everyone clocking in at open and standing around. And schedule by station: the line, the bar, the host stand and the floor each have their own breaking point, and one of them is usually the real bottleneck long before the others.
You can’t manage a schedule on gut feel once you’re past a single location. A few metrics tell you almost everything, and your POS should surface them without a spreadsheet. Labor cost as a percentage of sales is the headline. Sales per labor hour tells you how productive each scheduled hour is. Overtime is pure margin leakage — it usually means the base schedule is wrong, not that you’re busy. And the split between hourly wages and management salary shapes how much of your labor even flexes with volume in the first place.
| Metric | What it tells you | How to use it |
|---|---|---|
| Labor cost % | Total labor as a share of sales | Track against your target range daily, not just at month-end |
| Sales per labor hour | How much revenue each scheduled hour produces | Spot the dayparts where you’re over- or under-staffed |
| Overtime hours | Hours paid at a premium rate | Near zero is the goal — recurring OT means fix the base schedule |
| Scheduled vs actual | Whether shifts run long past their planned end | Chase the gap; unmanaged clock-outs quietly inflate cost |
The same discipline you’d apply to getting food cost under control applies here: measure it, set a target, and check it often enough to react while it still matters. Labor you review once a month is labor you’ve already lost.
Scheduling isn’t only a math problem — it’s a legal one, and the rules vary across the US and Canada. Overtime thresholds, mandated breaks and rest periods, and minor-labor restrictions all carry real penalties when you get them wrong. A growing number of North American cities also have predictive-scheduling (or “fair workweek”) laws that require posting schedules a set number of days ahead and paying a premium when you change a shift late. If you run more than one location, the rules can differ between them, which makes a single system that knows each location’s settings far safer than a shared spreadsheet nobody updates.
The reason scheduling stays painful in a lot of restaurants is that the tools don’t talk to each other. The schedule lives in one app, time clock in the POS, and payroll in a third system, so every week someone re-keys hours and hopes the numbers line up. That’s slow, and it’s where errors — and quiet overpayments — creep in. When scheduling, time tracking, sales and payroll sit in one platform, the loop closes: you build the schedule against live sales history, staff clock in on the same terminals that ring sales, actual labor cost shows up next to actual sales in real time, and approved hours flow straight to payroll with nothing re-typed. And because it’s offline-first, the time clock keeps running through an internet outage instead of leaving you guessing who worked what.
You don’t need a complicated process — you need a consistent one. Most well-run rooms follow roughly the same weekly loop, and it takes far less time once the data is doing the heavy lifting.
Novaryq puts scheduling, time tracking, sales and payroll preparation in one platform built for North American restaurants, so you build the schedule against real sales history, see labor cost against sales as it happens, and send approved hours to your payroll provider without re-keying. Overtime and break flags, minor-labor rules and per-location settings are handled where the schedule lives, and because the system is offline-first, the clock keeps running when the internet doesn’t. It’s multi-location native for groups that need each site on its own rules, and pricing is transparent and per-location with a month-to-month option. If you’re still deciding what a full platform should cost, our breakdown of what a restaurant POS actually costs lays out total cost of ownership.
It varies by format, but full-service restaurants often run labor in the range of 25 to 35 percent of sales, with counter-service and quick-serve typically lower. Treat it as a target range to manage against rather than a fixed number — set your own based on your concept and market, and check it daily instead of waiting for month-end.
Start from a sales forecast built on your POS history rather than copying last week. Break the day into dayparts, staff each to its own demand curve with staggered start and end times, and protect your peak revenue hours above all. Then track labor cost percentage and sales per labor hour so you can adjust before the week gets away from you.
Predictive-scheduling or “fair workweek” laws, now in effect in a number of North American cities, generally require employers to post schedules a set number of days in advance and to pay a premium when they change a shift on short notice. If you operate in one of these jurisdictions, post on time and treat late changes as a real cost, and use a system that knows each location’s rules.
It helps a lot. When scheduling, time tracking, sales and payroll share one platform, you build the schedule against live sales history, staff clock in on the same terminals that ring sales, labor cost shows up next to sales in real time, and approved hours flow to payroll with nothing re-keyed — which removes the manual re-entry where most errors and overpayments happen.