When labor cost is too high, the reflex is to cut hours. Sometimes that is right. Usually it is the fourth-best option, because the same money can be found in the hours you are already paying for — and those hours do not resign, complain or serve worse food.
📏 Measure output, not just cost
Labor percentage tells you what you spent. It does not tell you what you got. Two measures do:
- Sales per labor hour. Total sales divided by total hours worked. Track it weekly, by day, and compare against yourself. A Tuesday at half your Saturday's figure is telling you something specific about Tuesdays.
- Covers per labor hour, if your ticket sizes vary a lot. Same idea, measured in guests rather than money, which removes price changes from the picture.
The value is in the comparison, not the absolute number. Same restaurant, same shift, over time: if sales per labor hour is climbing, you are getting more from what you pay. If it is falling while sales are flat, hours have crept in somewhere — usually one station, one hour, one habit.
🔧 Where the wasted hours actually are
| Where | What it looks like | What to do |
| The dead hour | Full staffing at a time nobody comes | Staggered starts and staggered cuts. From the previous course |
| Prep | Done twice, or in quantities that get thrown away | Par levels, one prep list, one person owning it |
| Walking | Steps to the walk-in, the office, the far station, all shift | Move the thing, not the person. This is free and it is the most ignored fix in any kitchen |
| Waiting | Somebody blocked by somebody else's task | Reorder the sequence, or start one task earlier |
| Doing it twice | Rewriting a ticket, re-counting inventory, re-keying an order into the POS | The integrations from the previous courses exist for exactly this |
| Closing | Four people finishing at once because that is when the schedule ends | A closing list with names and times, so people leave as their part is done |
Pick one. Time it honestly for a week before and after. A task that runs twenty minutes shorter every day is more than a hundred hours a year, and nobody had to be told their hours were being cut.
🔁 Cross-training is a cost tool
A team where everybody does one thing needs more people to cover the same service, because coverage has to be bought per role. Cross-training changes the arithmetic:
- Fewer people cover more situations. A no-show at one station stops being a crisis.
- The shoulders get cheaper. One person who can prep, receive and run the counter replaces two half-shifts.
- It reduces overtime. More people are eligible to cover, so coverage stops landing on whoever is already at 38 hours.
- It is retention. People who learn things stay longer, which is topic 5 and cheaper than everything on this list.
Pay for it visibly — a small raise per station learned, or first choice of shifts. Cross-training that is only extra work with no reward is a thing people quietly avoid getting good at. How to actually run the training is the next book in the catalog.
🚫 The cuts that cost more than they save
- Cutting the closer so the opener arrives to yesterday's mess and spends two hours on it. Same hours, worse start.
- Cutting prep so the line improvises during service. This shows up as slower tickets and food cost, not as labor.
- Cutting the busser so servers clear tables and stop selling. The saving is real; the lost turn is bigger.
- Cutting the trainer so the new hire learns from whoever is nearest. The previous course explained what that costs.
- Sending people home ten minutes after they arrive. The saving is minutes and the message lasts months.
- Cutting the same person every time. That is not a labor decision, it is a resignation with a delay on it.