Sizing a shift

Ask how many servers a restaurant needs and you will get ratios — tables per server, covers per cook — offered with confidence and no context. They are a reasonable starting guess and a poor operating rule, because they encode someone else's menu, room and service style.

Your own numbers are better, and you already have them.

Start with demand by daypart

Pull covers by hour across several representative weeks, including a bad-weather day and a holiday, and look at the shape rather than the total. Almost every restaurant discovers the same thing: the day is not one thing. A room that averages steady demand is usually two peaks and a long trough, and a flat schedule serves none of them well.

Schedule to the shape. That means staggered starts rather than everyone arriving at once, and it means being honest about which hours genuinely need the full team.

Use sales per labor hour to check yourself

Sales per labor hour — net sales divided by hours worked — answers the question a labor-cost percentage hides: were the people who were on actually busy? Track it by daypart, because a strong dinner will mask an overstaffed afternoon indefinitely if you only look at the day.

The useful comparison is your own history, not a published benchmark. A daypart whose figure is drifting down while covers hold steady is a staffing question with a clear answer.

What changes the number

• ✓Menu complexity. Longer preparation moves the constraint into the kitchen, where more front-of-house does not help.
• ✓Service style. Coursed dining needs more attention per cover than counter service, and the ratio reflects that.
• ✓Room layout. A long walk from the pass to the far section costs a measurable share of every server's shift.
• ✓Ordering channels. Kiosk and QR ordering shift work from taking orders to running food and expediting, which changes the mix of roles rather than the headcount.
• ✓Table turn speed. Faster turns mean more covers through the same seats, which needs more hands even though the room is no bigger.

Then check it against what happened

A staffing model is a hypothesis, and the test is comparing what you scheduled against what was worked and what was sold. If a shift consistently runs long, it was under-scheduled regardless of what the ratio said; if it is quiet for two hours in the middle, the shape is wrong even when the total is right.

Doing this without a spreadsheet

Opero puts labor cost beside the sales it produced, and holds the published schedule and clocked hours in the same system, so scheduled against actual is readable rather than reconstructed. That is what turns staffing from an annual argument into a weekly adjustment.

One thing not to expect: Opero does not forecast demand or draft schedules for you. It shows you what happened accurately, and the sizing judgement stays yours. Scheduling, the time clock and labor cost reporting are on the Growth plan and above.