To build a restaurant staff schedule that actually works, forecast covers first and staff to that number, not to last week’s rota or gut feel. Restaurants that schedule against forecast demand cut both overtime and understaffed peak shifts, the two most expensive scheduling mistakes in the business.
Most independent restaurants build next week’s schedule by copying last week’s and tweaking it for known days off. That approach staffs for an average night, not for the actual mix of quiet Tuesdays and packed Saturdays your reservation book already tells you about. The fix isn’t more staff. It’s matching the staff you have to the demand you can already see coming.
Key takeaways
- Main solution: Forecast covers by day and hour from booking data, then build the rota to match
- Expected result: Lower overtime and labor cost with fewer understaffed peak shifts
- Time to implement: 2-3 hours weekly once the process is set up
- Cost: Free (uses data you likely already have)
Before you start
Scheduling well requires knowing your demand before you know your staffing.
What you’ll need:
- Reservation and covers data by day and hour, ideally 4+ weeks of history
- A list of roles and their typical staff-to-cover ratios
- Your labor cost target as a percentage of sales
- A record of recent call-outs and no-shows by shift
Know your baseline: Pull last month’s actual covers by hour and compare them to who was scheduled. Gaps between the two, overstaffed slow Tuesdays, understaffed Saturday peaks, are exactly what forecast-based scheduling fixes.
Step 1: Forecast covers before you schedule anyone
Every schedule should start with a number: how many guests are you serving, and when.
What to do:
- Pull reservation and walk-in covers by day and hour for the past 4-8 weeks
- Identify your recurring pattern (Tuesday is always light, Friday always peaks 7-9pm)
- Adjust for known events, holidays, or seasonal shifts
- Produce a covers forecast for the week you’re scheduling
Why forecast covers instead of scheduling by feel:
| Approach | Result |
|---|---|
| Copy last week’s schedule | Misses demand shifts, drifts from reality over time |
| Schedule by manager instinct | Inconsistent, depends on who’s building the rota |
| Forecast covers, then staff to match | Matches labor to actual demand, adjusts automatically as bookings change |
Step 2: Convert covers to staffing levels by role
Once you know expected covers, translate that into headcount for each role.
What to do:
- Set a staff-to-cover ratio for each role based on your service style
- Apply the ratio to your hourly forecast
- Round up for safety during your top 2-3 peak hours
- Adjust for skill mix, not just headcount
Typical staff-to-cover ratios by role:
| Role | Peak ratio | Off-peak ratio |
|---|---|---|
| Servers | 1 per 15-20 covers | 1 per 25-30 covers |
| Bussers | 1 per 40-50 covers | 1 per 60+ covers |
| Line cooks | Based on menu complexity and station count | Reduced stations |
| Host | 1-2 | 1 |
| Bartender | 1 per bar section | 1 |
Worked example: A restaurant forecasting 90 covers between 7-8:30pm needs roughly 5-6 servers at a 1:15-18 ratio, versus 3 servers for a 45-cover Tuesday night at the same time. Scheduling both nights identically wastes labor on Tuesday and understaffs Friday.
Step 3: Build the rota against the forecast
With headcount targets by hour, build the actual schedule.
What to do:
- Schedule your strongest staff against your highest-forecast shifts
- Stagger start and end times to match the demand curve, not fixed shift blocks
- Build in overlap at shift changes so nobody works the peak alone
- Confirm total scheduled hours against your labor cost target
Stagger starts instead of one fixed call time:
| Poor practice | Better practice |
|---|---|
| Everyone starts at 5pm regardless of forecast | Early servers start at 4:30 for setup, mid-shift servers start at 6 ahead of the 7pm peak |
| Same headcount 5pm-close | Headcount ramps up before peak, tapers down after |
Step 4: Post the schedule early and set a shift swap policy
Advance notice reduces last-minute scrambling for both you and your team.
What to do:
- Post schedules at least 7 days ahead; 14 days if your city or state requires it
- Define a clear shift swap process
- Require manager approval on swaps to protect skill mix per shift
- Track who initiates frequent swaps or call-outs
Why advance notice matters: unpredictable scheduling is one of the most commonly cited reasons hourly restaurant workers give for quitting, and several jurisdictions now require advance notice by law. Oregon has a statewide predictive scheduling law, and Chicago, Seattle, Philadelphia, and New York City each enforce local fair workweek ordinances requiring advance schedule posting and premium pay for late changes (Oregon Bureau of Labor and Industries; city ordinance summaries via Littler). These laws generally apply only to larger operators, Oregon’s covers employers with 500 or more employees worldwide, and the city ordinances carry similar size or location-count thresholds, so most independent single-location restaurants fall outside them today. Check your specific city and state requirements regardless, since thresholds shift and more jurisdictions are considering similar rules.
Shift swap policy template:
| Rule | Why |
|---|---|
| Staff find their own qualified swap first | Reduces manager workload |
| Manager approves before the swap is final | Protects skill and experience mix |
| No swaps within 24 hours of the shift without manager sign-off | Prevents last-minute gaps |
| Track swap frequency per employee | Flags scheduling or engagement issues early |
Step 5: Handle the double shift question deliberately
Doubles (working two shifts in one day) are sometimes the fastest way to cover a gap, but they carry real costs if overused. See the full breakdown in what is a double shift for the labor-cost, fatigue, and overtime tradeoffs before defaulting to doubles as your go-to fix.
What to do instead of defaulting to doubles:
- Use doubles for genuine gaps (call-outs, events), not as standing policy
- Offer doubles rather than mandate them
- Track how often the same people are picking up doubles
- If doubles happen every week, treat it as an understaffing signal, not a scheduling win
Common mistakes to avoid
Scheduling from last week’s rota instead of this week’s forecast
Demand shifts gradually, and a rota copied week over week drifts further from reality each cycle. Re-forecast every week, even if it feels repetitive.
Ignoring the gap between forecast and actual until payroll
If you only compare scheduled hours to actual demand after the fact, you’ve already paid for the mismatch. Check the forecast against the schedule before the week starts, not after.
Treating all peak hours the same
A Friday 7pm peak and a Tuesday 7pm peak are not the same size. Apply your ratios to the actual hourly forecast for that specific day, not a flat “dinner rush” assumption.
No shift swap process
Without a defined process, staff either can’t get coverage when they need it or swap shifts without you knowing who’s actually on the floor. Both create risk.
Using doubles as a permanent fix for understaffing
A double shift solves tonight’s gap. It doesn’t solve the reason you’re short-staffed. If doubles are constant, the real fix is hiring or rebalancing your ratios.
How to measure success
Track these weekly for the first month:
| Metric | Before | Target | How to track |
|---|---|---|---|
| Overtime hours | Your baseline | -20%+ | Payroll/time-tracking export |
| Labor cost as % of sales | Your baseline | Within your target range | Labor $ / sales $ |
| Understaffed peak shifts (guest complaints, slow tickets) | Your baseline | Fewer per week | Manager log |
| Last-minute call-outs | Your baseline | Fewer per month | Scheduling log |
Tools that help
Forecast-based scheduling depends on seeing your booking data clearly, not just building the rota itself.
Booking data and reports show covers by day and hour, the raw input every forecast needs. Without this, scheduling is still a guess, just a more organized one.
Reservation pacing limits how many covers can book into a single time slot, which keeps your forecast (and therefore your schedule) more stable week to week.
Historical trend views let you compare this month’s Friday pattern to last month’s, catching demand shifts before they force a reactive schedule change.
If your current system doesn’t show covers by hour clearly, Resos includes reporting that forecasts covers from your booking data before you build the rota.
Frequently Asked Questions
How do I create a restaurant staff schedule?
What is a good server-to-table or server-to-cover ratio?
How far in advance should I post the schedule?
What's a fair shift swap policy?
How do I reduce scheduling-related turnover?
The bottom line
Good restaurant staff scheduling starts with a number, not a habit. Forecast covers by day and hour from your own booking data, convert that to staffing levels by role, then build the rota to match. Post it early, give staff a real shift swap process, and use doubles sparingly and deliberately rather than as your default fix.
The restaurants that treat scheduling as a data problem, not a weekly chore, consistently run leaner labor costs and steadier service than the ones still copying last week’s rota.
Related guides: Restaurant manager duties and SOPs | Capacity planning | RevPASH