Chapter 19 — Key Takeaways

Labor Management: Scheduling, Labor Cost Control, Overtime, and the Balancing Act of Staffing

The core claims

1. The schedule is a financial document. It is the only one in the building you write before the money moves. Bellwether's labor line is \$500,000 across fifty-two schedules — roughly \$9,600 of commitment per schedule, typically written in forty minutes.

2. Labor is the half of prime cost you can move this week. Food cost changes over a purchasing cycle. Labor changes on Tuesday at nine o'clock. That responsiveness makes it the most controllable line in the business and the most frequently mismanaged.

3. "Labor" is four things, and most operators quote one of them. Wages, employer payroll taxes, workers' compensation, benefits. At Bellwether the last three total \$73,920 — 4.8 points of sales. An operator quoting wages-only labor is understating by nearly five points.

4. Name your denominator, in writing, every time. The same \$10,970 of weekly labor reads as 40.4%, 36.8%, or 36.0% depending on which week you divide by. Half the arguments about labor cost are two people using different denominators and neither one saying so.

5. The fixed floor is bigger than it looks. Bellwether's is \$191,895 — 12.4% of sales — committed before a single hourly hour is scheduled. It is why slow services are dangerous and why the shape of your week matters more than your average.

6. A high labor percentage on a service is not a reason to close it. Sunday brunch runs 48.5% labor and contributes \$64,584 a year toward fixed cost. Closing it would make the annual labor percentage worse by 1.1 points. Measure contribution, not percentage.

7. Staffing is a staircase, not a slope. Covers between 85 and 110 cost about 0.36 labor hours each. Covers between 110 and 130 cost about 1.38. Know where your risers are, and work to fill the step you are standing on rather than climbing to the next one.

8. Build the schedule first; let the percentage be the output. A percentage-of-sales labor model funds the most hours in the quarter with the fewest guests. Ask any operator with one: how many hours a week does that fund? If they can't answer in ten seconds, the number isn't connected to a schedule.

9. Cut the end of a shift, never the beginning. A cook who starts thirty minutes late is thirty minutes behind at the moment the first ticket lands, and never catches up. You save \$10.50 of wages and lose two second seatings.

10. A labor report measures dollars paid — not work done, and not damage sustained. The second Friday in October cost \$116.70 (1.7% of the night's sales) and the restaurant came out **\$12 ahead on wages.** The fifteen-hour sous, eleven tables past standard, and a cook alone on two stations for four hours appear nowhere.

11. The forecast is the failure more often than the crisis is. That Friday was written to a 120-cover band with 142 on the books. It was 9.5 hours short — about \$130 all-in — before anybody failed to show up.

12. A labor percentage that improves while ticket times, comps, turnover, and reviews worsen is not an improvement. It is a transfer, from the guest experience and your staff into the labor line, at a very poor exchange rate.

The key formulas

Labor cost %   = (wages + payroll taxes + workers' comp + benefits) ÷ net sales

SPLH           = net sales ÷ labor hours          (productivity; moves with check average)
CPLH           = covers ÷ labor hours             (production; comparable across concepts)

All-in hourly cost = wage × (1 + payroll tax rate + workers' comp rate)
Overtime hour      = wage × 1.5, then grossed up the same way

Rules of thumb (ranges, not laws — build your own)

Bellwether, at a glance

The verdict: 30.2% is not reachable by writing a tighter schedule. Scheduling discipline is worth about 1.2 points. Revenue above plan is worth about 2.3 points. The last 2.3 points must come from the kitchen's production model, from the owners' compensation, or from sales roughly 17% above plan — and a plan that shows 30.2% without naming which one is doing the work is an arithmetic identity solved for the answer somebody wanted.

Key terms

labor cost percentage · fixed labor · variable labor · sales per labor hour (SPLH) · covers per labor hour (CPLH) · labor forecasting · staffing guide / labor matrix · overtime · split shift · cutting the floor

What you should be able to do Monday morning

Take your last four weeks of point-of-sale data and build a staffing guide in cover bands for one service — positions and hours for each band, with the riser thresholds marked. Then write next week's schedule to it, post the forecast on it, add the hours column and check every person against 40, and write the cut order before service. On Monday, read the variance in hours first, dollars second, percentage last — and put ticket times, comps, review scores, and departures on the same page as the labor number, so that a "win" and a slow degradation cannot look alike.