How to build a restaurant shift roster by meal period

Start with arrivals by meal period

A dining-room total hides when the work arrives. Record reservations, walk-ins, table sizes and expected arrival times in 15-minute intervals. Turn those into likely tickets at the pass, drinks at the service station, table resets and payments. A party sitting down together does not necessarily place its whole order immediately; measure the time between arrival and kitchen release.

The retained format is a 60-seat restaurant without alcohol. Its mature demand is 90 covers per trading day. Our proposed lunch/dinner split gives 540 weekly covers and 2,340 in an average month using 52 weeks divided by 12, matching the retained model’s mature volume. The opening ramp is lower. Neither the split nor the quarter-hour peaks came from reservations or a timed trial. Example boundary

For a trial, keep the arrival sequence rather than replacing it with a flat hourly average. In this worksheet the dinner peak occurs before 20:00, when both paid servers are present. A later peak would invalidate the handoff to one employee and the owner.

Add prep, reset and close to the customer windows

Use a position sheet before assigning people. Give each station a start time, handoff and last task. Preparation, receipt checks and cleaning consume paid time even when the doors are closed. The template below keeps two cooks during both meal windows, staggers the dinner handoff at 18:00 and pays the close through 22:00. The dining close finishes at 21:30; dish and kitchen closing continue afterward.

The daily template uses 33.0 primary and 16.5 support hours. Across 6 trading days, add 2 primary hours for Monday count/ordering and 1 support hour for Tuesday receiving. That reaches 200 and 100 respectively. This allocation interprets the two budget pay groups; the base model itself does not assign them to these jobs.

The 15:00–16:00 dish gap is deliberate: the opener must be able to handle the remaining reset within that paid interval. Confirm the tasks and wash volume; if they do not fit, the template needs another paid shift or a different handoff. A scheduled interval is not a measured task completion.

Check the workweek by person, not only by position

Assign the positions to actual available people and then total each person’s week. A shared pool of hours cannot tell you whether one cook has been double-booked or whether the Sunday opener exists. The slot IDs below are hypothetical adults. They describe a schedulable allocation, not the restaurant’s established headcount or a claim that this many part-time hires are available.

K5 covers the weekend opener and the Tuesday–Thursday lunch shifts; those jobs do not overlap. H1’s lunch and evening host work produces split days, which need local-rule and availability review. K1’s count shift is on the closed day and is paid. The greatest employee total is 34.0 hours here; the aggregate budget alone would not reveal that total.

Keep one fixed workweek for payroll review. Federal guidance for covered nonexempt employees requires overtime above 40 hours in the workweek at at least 1.5 times the regular rate; weeks cannot be averaged together. The worksheet assumes one straight-time rate per employee and no additional regular-rate components. Check the actual jurisdiction, duties, pay terms and scheduling obligations before implementing it. Federal workweek guidance

Make owner work and relief visible

The owner is already paid in the retained case: $5,500 gross per month, or $6,325.00 with the same assumed loading. Owner labor must have an available interval and the skill to cover the position. It cannot simultaneously host the lunch queue, relieve a cook and approve a delivery.

The owner takes the host position from 13:30 and the second server position from 20:00. Relief is reserved separately, so those guest-service handoffs are not also counted as break cover. Keep a dated break sheet showing which qualified person covers which station. The relief blocks are a capacity allowance to test; they do not establish that all local break requirements can be met.

Under federal hours-worked guidance, permitted work still counts even when it happens after the scheduled finish. Short rest breaks count as work time, and a meal can be unpaid only when the employee is fully relieved of duties. This cost worksheet pays the full elapsed intervals; it makes no automatic meal deduction. Record actual release and return times and revise coverage if the real policy changes the usable station hours. Hours-worked guidance

Reconcile the roster to the existing payroll budget

Keep gross wage cost, payroll loading and productive station time as different measures. A paid break remains in wage cost while the station needs relief. The case’s hourly wages and loading are explicit assumptions, not wage quotes or legal minimums.

Loaded employee cost averages $30,896.67 per month using 52 divided by 12, matching the primary-plus-support payroll already in the case. Working-owner compensation remains separate. Do not add these employee wages to the per-cover variable costs again: the model treats this roster as a fixed commitment.

This reconciliation proves that the proposed slots use the stated hours and wage assumptions. It does not prove service quality, demand, staff availability or local compliance. Read the restaurant unit-economics guide when linking a measured service trial back to the monthly forecast.

Price an extra shift before solving a coverage gap

Suppose an additional 8-hour preparation shift is needed beyond the weekly template. Giving it to K1 raises that slot from 34.0 to 42 hours. With this example’s one regular rate, 2 hours cross the federal threshold. The extra gross wage is $176.00 for all added hours plus $22.00 in overtime premium, totaling $198.00.

Applying the same assumed loading produces $227.70. A qualified relief worker who stays below the threshold would cost $202.40 at the same rate, a $25.30 difference. The comparison concerns genuinely additional work; replacing an already budgeted shift requires a separate net-cost bridge. Neither option is added to the published model by this article.

Before buying extra hours, check the cause: a delivery overlapping lunch, too much prep on the opener, tickets concentrated in a late dinner wave, or closing work repeatedly continuing after clock-out. Move work only when food processes, employee availability and premises constraints allow it. A lower sales forecast does not automatically remove a mandatory station.

Release the roster only after a full mixed-service trial

Test a complete lunch and dinner with the actual menu and table pattern, including replenishment and cleanup. Save failures as well as successful runs. If the trial needs more paid coverage, revise the operating assumptions before claiming that the financial case supports the expanded promise. The retained case has negative first-year EBITDA and does not recover its project investment within the base 60-month horizon; a roster worksheet does not reverse those results. Use the profitability guide to judge the financial effect.