Good restaurant staff scheduling comes down to one habit: forecast how busy each part of the day will be, then put labor hours where the sales actually are. Match your schedule to demand daypart by daypart, watch sales per labor hour as you build it, and you can hold labor cost inside your target range without leaving the floor short when it counts.
Most managers still schedule the way they were taught: copy last week, tweak for who called out, post it Thursday, hope. That gets you a schedule that’s fair to the staff and roughly wrong for the business. Some shifts you’re paying three people to lean on the pass; others you’re 86ing items because the line is buried. Both cost you money. Scheduling to demand fixes both at once, and it doesn’t take a data science degree, just a little math and the discipline to look at the numbers before you build.
Start with covers, not gut feel
Before you touch the schedule, forecast the week. Pull the last four to six weeks of sales out of your POS and break each day into dayparts, not just a single daily total. A Friday isn’t one shift; it’s a sleepy lunch, a dead mid-afternoon, and a dinner rush that does half your day’s sales in three hours. If you schedule to the daily average, you overstaff the quiet stretches and understaff the peak.
Look at covers (guests served) or transactions per daypart, then layer in what you know that the POS doesn’t: a local event, a holiday, weather, a patio that only fills when it’s above 70. Your forecast doesn’t need to be perfect. Getting each daypart within roughly 10 to 15 percent is enough to schedule intelligently, and your accuracy climbs fast once you’re checking forecast against actual every week.
Match labor to your peaks
Once you know when the guests show up, staff to that curve instead of to round-number shifts. The mistake is scheduling everyone 4-to-close when your dinner volume doesn’t build until 6. Stagger it. Bring your first server in for the early tables, add bodies as the forecast climbs, and cut back toward close so you’re not paying a full crew to reset the dining room for fifteen stragglers.
Two moves do most of the work here:
- Stagger start and end times in 30- to 60-minute steps that track the build and the fade, rather than three people all clocking in at 5:00.
- Use cut times, not just start times. Write “5:00–close, cut at your discretion after 9:00” into the plan so a slow Tuesday actually saves you money instead of running four servers till midnight out of habit.
Keep your strongest people on the peak and lighter, cheaper coverage on the shoulders. A closer who can turn tables fast during the rush is worth more at 7 p.m. than at 10.
Use sales per labor hour as your guardrail
Labor cost percentage tells you how you did after the week is over. Sales per labor hour (SPLH) tells you whether a shift is staffed right while you’re building it. The math is simple: forecast sales for a daypart divided by the labor hours you’ve scheduled into it.
Set a target band for each daypart. Full-service dinner might run healthy at $55 to $70 in sales per labor hour; a counter-service lunch might sit higher. If a daypart pencils out well below your band, you’ve scheduled too many hours for the volume, cut one. If it’s way above the band, you’re about to run people ragged and blow up service, add coverage. SPLH turns “this feels about right” into a number you can defend to your GM and adjust before the shift, not after.
Watch out: SPLH is a guardrail, not a ceiling. Chasing a sky-high number by cutting to the bone tanks service, turns tables slower, and quietly kills your check average and tips. The goal is the band, not the maximum. A shift that’s too lean costs you in ways the labor line never shows.
A worked example: tying the schedule to a labor cost percentage
Take a mid-size full-service spot forecasting $4,500 in sales this Friday, with a blended hourly wage around $16. Here’s what a demand-matched schedule looks like when you lay it out by daypart:
| Daypart | Forecast sales | Labor hrs | Labor $ | SPLH |
|---|---|---|---|---|
| Open / prep (9–11) | — | 8 | $128 | — |
| Lunch (11–2) | $1,350 | 22 | $352 | $61 |
| Afternoon lull (2–5) | $450 | 10 | $160 | $45 |
| Dinner (5–9) | $2,450 | 34 | $544 | $72 |
| Close (9–11) | $250 | 10 | $160 | $25 |
| Total | $4,500 | 84 | $1,344 | $54 |
That schedule lands labor at $1,344 on $4,500 in sales, or about 29.9 percent, right inside a typical 28–32 percent target for full-service. Notice where the story is: dinner is doing $72 in sales per hour and earning its coverage, but the close is scraping $25. That’s your signal to trim one closer or send someone home at 10 instead of 11. Pull four labor hours off the close and you drop to roughly $1,280, about 28.4 percent, without touching the rush that actually makes your night. Run those same numbers across the week through a restaurant labor cost calculator and you’ll see fast which shifts are quietly dragging your percentage up.
Avoid the twin traps of over- and understaffing
Overstaffing is the quiet killer because nobody complains. Guests are happy, staff are relaxed, and you don’t feel the bleed until the P&L shows labor at 36 percent and your profit is gone. Understaffing screams louder, slower tickets, walkouts, a one-star review about the wait, and it costs you on the revenue side even though the labor line looks great that week. A demand-matched schedule with SPLH guardrails is how you stay out of both ditches.
A few operating rules that keep the plan honest:
- Compare forecast to actual every week. Ten minutes reconciling last week’s schedule against what really happened is the single fastest way to make next week’s forecast sharper.
- Protect the peak, flex the shoulders. Never cut into your busiest 90 minutes to save labor. Find the savings in prep, the lull, and the close.
- Build in cut authority. A schedule that can shrink when it’s slow beats a rigid one every time.
- Know your fixed floor. Some hours can’t be cut, one cook and one server minimum, whatever the volume. Schedule those first, flex everything above them.
Do this for a month and the payoff compounds: tighter forecasts, calmer shifts, and a labor percentage you’re steering instead of explaining after the fact.
Frequently asked questions
What is a good labor cost percentage for a restaurant?
It depends on your model. Full-service restaurants often target 28–32 percent of sales, while quick-service and counter-service can run lower because they use fewer labor hours per dollar. Fine dining frequently runs higher. Set your own target from your P&L rather than a generic benchmark, then schedule to defend it.
How do I forecast covers for scheduling?
Pull four to six weeks of POS data, break each day into dayparts, and average the covers or transactions for each. Then adjust for anything the data can’t see, events, holidays, weather, and seasonality. Getting within 10–15 percent per daypart is enough to schedule well, and your accuracy improves as you compare forecast to actual each week.
What is sales per labor hour and how do I use it?
Sales per labor hour (SPLH) is forecast sales for a period divided by the labor hours scheduled into it. Set a target band per daypart and use it as a guardrail while you build the schedule: below the band means you’re overstaffed, well above it means you’re likely understaffed and about to strain service.
How far ahead should I post the schedule?
Post at least a week out, and more if predictive-scheduling laws apply in your city, some require 10 to 14 days’ notice and pay penalties for last-minute changes. Earlier posting also cuts no-shows and swap requests, which protects the coverage your forecast is counting on.
Should I cut staff early when it’s slow?
Yes, if service holds. Build cut times into the schedule and give shift leads authority to send people home once volume drops below plan. Just protect your peak and your fixed minimum coverage, cutting too deep slows tables and costs you more in lost sales and tips than you save on wages.
How does scheduling connect to my overall bookkeeping?
Labor is usually your largest controllable cost, so your schedule and your books have to agree. When payroll, sales, and hours flow into clean bookkeeping, you can see labor percentage by week or daypart and catch drift early. Checking a schedule against a restaurant labor cost calculator before you post it keeps the plan and the P&L pointed the same direction.
Great scheduling only pays off when your numbers are clean enough to act on. Tabby is AI bookkeeping built for small businesses and the self-employed, so restaurant operators can see labor, sales, and margins clearly without drowning in spreadsheets. Start a free trial and spend less time reconciling and more time running the floor.
