How to Reduce Labor Cost Without Cutting Hours

When labor cost runs hot, the first thing most operators do is cut hours. Send people home early, trim a shift, fire someone. And almost every time, that move backfires inside 2-3 weeks, because the operation actually needed those hours. It just needed them at different times of the day. So the fix isn’t fewer hours. It’s changing when you have them scheduled.

Run an SPLH audit before touching the schedule #

Sales per labor hour (SPLH) is total sales divided by total labor hours for a period. Higher is better. It tells you whether the labor you scheduled actually produced enough revenue to be worth what you paid for it.

SPLH = Total Sales ÷ Total Labor Hours

Target SPLH by segment:

• QSR / fast food: $80-120
• Fast casual: $60-90
• Casual full-service: $50-80
• Fine dining: $80-150
• Coffee shop: $60-90
• Bar (no kitchen): $100-200
• Pizza (delivery + takeout): $90-130

Pull last 4 weeks of SPLH by shift and find the worst ones. That’s where the labor leak is. Here’s the part people get backwards: cutting hours on shifts that are already tight just wrecks your service. The shifts you want to go after are the low-SPLH ones, because that’s where you can fix the leak without anybody on the floor feeling it.

For full SPLH math, see Sales Per Labor Hour: The Number That Tells You Who To Cut First.

Fix 1: Stagger shift starts #

The biggest single labor leak in most operations is dumb simple: everyone clocks in at 4pm for a 5:30pm dinner rush.

Watch how it actually plays out. Kitchen comes in at 4. Prep starts at 4. Servers come in at 4, side work starts at 4, bar comes in at 4. From 4:00 to 5:00 the whole crew is grinding through prep that could have started earlier. From 5:00 to 5:30 everybody’s finishing prep and getting set. Then the first guest walks in at 5:30 and only now is the team actually working at full capacity.

That hour of “everyone at 4” is mostly paid standing-around. On a 6-person team at $20/hour fully loaded, that’s $120/night of paid prep time that barely touched the register.

The fix: stagger starts based on what each role needs.

• Kitchen prep cook: 3:00pm (gets the prep done in the kitchen’s own quiet space)
• Line cooks: 4:30pm (after prep is staged)
• Server side work: 4:30pm
• Servers (taking tables): 5:00pm
• Bartender: 4:30pm
• Host: 5:00pm
• Manager: 4:00pm (sets up, runs preshift, oversees)

Your total labor for the night barely moves. What changes is how much of that labor is pointed at a paying customer instead of a prep table. The floor starts making money the minute it opens, because the prep is already done before the first guest sits down.

Typical savings: 2-4 labor hours per night = $40-80 per night = $2,000-4,000 per month.

Fix 2: Cross-train so one server can run food, host can bus #

Cross-training is the lever everyone skips, because it costs you training time up front and the payoff isn’t tonight. But once it’s done, one person can cover two or three roles on a slow night instead of you carrying bodies you don’t need.

Front-of-house cross-training:

• Servers learn host duties (greeting, seating)
• Servers learn busser duties (clearing, resetting)
• Bartenders learn food running
• Hosts learn busser duties

Back-of-house cross-training:

• Line cooks learn each station (one cook can cover sauté + grill on a slow shift)
• Prep cooks learn line stations for cover
• Dishwasher learns light prep

When a shift comes in slower than you forecast, cross-trained people cover the gaps instead of standing around watching each other. A shift you scheduled for 6 can run on 4 when it’s slow, with everybody picking up a second job.

Typical savings: 15-25% of labor on slower shifts through coverage flexibility.

The investment: 2-4 weeks of paid training time per cross-train (server learns bar, etc.). The payback period: 60-90 days at average volume.

Fix 3: Forecast-based scheduling #

Most independent operators schedule by gut. “Last Saturday we ran Y people, so this Saturday we’ll run Y again.” That’s fine when business holds steady. It falls apart the second business moves, which is basically always.

Here’s the discipline:

1. Pull last 4-6 weeks of same-day-of-week sales.
2. Calculate weighted average (weight recent weeks higher).
3. Apply seasonal adjustment.
4. Set labor target as percentage of forecasted sales.
5. Schedule against the labor budget, not against last week’s headcount.

Worked example: Saturday night forecast of $9,400 in sales. Labor target 28%. Labor budget: $2,632. At $19/hour fully loaded average, that is 138 labor hours. Build the schedule against 138 hours, not against “last Saturday we used 145.”

This catches over-scheduling before it lands on a printed schedule, instead of after, when you’re staring at a labor number you can’t fix. Operations that switch to forecast-based scheduling usually pull labor cost down 2-3 points inside 60 days, and service doesn’t suffer for it.

Fix 4: Kill overtime before it lands #

Overtime costs 1.5x base wage. A line cook earning $20 base ($25 fully loaded) costs $30 base in overtime ($37.50 fully loaded). Six hours of overtime a week is $225 in extra labor for output you’d have gotten at straight time anyway. You’re paying time-and-a-half for nothing.

And it almost always happens the same way: nobody’s watching running hours through the week. Someone quietly hits 40 by Friday afternoon, so any Friday-night or Saturday shift they touch is overtime, and you don’t find out until you run payroll.

The fix is real-time OT tracking. Most modern POS and time-clock systems will flag an employee creeping up on 40 hours, which gives the manager a chance to shuffle the back half of the week and stay under, or to look at it and decide the OT is actually worth it. Both are fine. Flying blind isn’t.

Common over-spent OT scenarios:

• Saturday shift that pushes an employee from 36 to 44 hours (8 hours of OT)
• Picking up a coworker’s call-out shift without checking the running total
• Salaried managers asking hourly staff to “stay another hour” without thinking about the wage rate

Typical savings: $300-1,200/month depending on operation size, just from tracking and avoiding rather than running hot.

Fix 5: Cut labor through retention #

Here’s the math almost nobody runs: turnover costs you way more than over-staffing ever will.

A single hourly quit runs $5,864 (HigherMe Restaurant Turnover Report 2024). Stop one quit a quarter and that’s $23,456 saved over the year. That number swallows the typical $5-10K you’d save from all the over-scheduling fixes above combined. This is the one I’d chase first.

Retention spend that consistently pays:

• Schedule predictability: Post schedules 2 weeks in advance, honor them.
• Manager training: Manager-attributed quits run 55-70% of total (Gallup). Better managers = lower turnover.
• 30/60/90 day check-ins: Most quits happen in first 90 days. Structured check-ins catch problems before they become exits.
• Cross-training: Lets employees pick up extra hours when they want, reduces feeling of being stuck in one role.

For full retention math, see What Restaurant Turnover Actually Costs.

The 5 KPIs that matter weekly #

Track these weekly:

1. Labor cost percentage (fully loaded labor ÷ sales × 100). Target: segment median minus 4-6 points.
2. SPLH by shift. Target: segment-typical for each daypart.
3. OT hours as % of total labor hours. Target: <2%.
4. Turnover rate (annualized). Target: industry minus 15-25 points.
5. Schedule adherence: how many shifts started/ended on the scheduled time vs adjusted. Target: 90%+.

The dashboard for all five is one spreadsheet you update Monday morning. Nothing fancy. The operators who watch these five every week run circles around the ones staring at a single aggregate labor cost percentage and wondering why it won’t budge.

What this looks like in the calculator #

The labor cost calculator on this site runs both labor cost % and SPLH for you. Pair it with the prime cost calculator to see the whole picture against benchmarks. Whatever you come in under target is your budget for the retention stuff that actually keeps people.

What to do today #

Pull last week’s shifts by daypart and calculate SPLH for each one. Find the bottom 3. That’s your leak. Now look at when each role clocks in on those shifts. If everybody’s punching in at the same time, the staggered-start fix by itself closes most of the gap, and you haven’t sent a single person home.

None of this is about cutting people. It’s about putting them on the clock at the right times. Those two are not the same thing, no matter how the labor number makes you feel on a Monday.

Sources: 7shifts, Toast, HigherMe Restaurant Turnover Report, NRA 2024 Industry Factbook, Restaurant365.