Everything in this course is worth nothing without a fixed fifteen minutes in the week when somebody looks. Not because the numbers are complicated, but because labor is decided weekly, and a monthly review tells you about four weeks you can no longer change.
🗓️ The weekly fifteen minutes
- Actual sales against forecast. By day. How wrong were you, and in which direction?
- Actual hours against budget. Also by day, because a good total can hide a bad Tuesday and a short Saturday.
- Labor percentage, split between kitchen and floor.
- Overtime. Who, how much, and why. Every hour of it should have a reason you can say out loud.
- One thing to change next week. One. A review that produces five actions produces none.
Do it on the same day every week, right after you approve hours. Fifteen minutes, with the previous course's four numbers open beside these — they are the same week seen from two sides.
📉 Reading the variance
| What you see | Usually means | Where to look |
| Hours on budget, percentage high | Sales missed, not overspending | The forecast, and why it was optimistic |
| Hours over, percentage fine | A busy week absorbed it — this time | Whether the extra hours were decided or just happened |
| Kitchen fine, floor high | Too many on the floor at the shoulders | Staggered starts and cuts, from the previous course |
| Overtime creeping every week | You are short a person and paying a premium to avoid noticing | Headcount, and the swap approvals |
| Scheduled and paid hours drifting apart | Early clock-ins, late clock-outs, added shifts | One station and one hour. It nearly always is |
| Everything fine, service complaints rising | You have cut into capacity, not waste | Sales per labor hour, and topic 4 |
The last row is the one to take seriously. A labor percentage that improves while reviews get worse is not an achievement, it is a trade you made without deciding to.
🍳 Prime cost, monthly
Once a month, put food and labor together as a percentage of sales, because they trade against each other and neither one alone can tell you whether you are winning:
- Labor down, food up? You bought convenience — pre-cut, pre-portioned, pre-made. Fine, if the total improved. Check that it did.
- Food down, labor up? You are prepping more from scratch. Same question, same answer.
- Both drifting up? Menu prices have not moved in a while, and everything else has.
- Prime cost stable while sales grow? That is the shape you want, and it is the point where the fixed part of your labor starts working for you.
Track it monthly on one line, next to last month and the same month last year. Most restaurants that fail slowly do so with a prime cost that moved two points a year for three years, which is invisible at any single review and obvious on one page.
🎯 What good looks like, and what comes next
You do not need to hit a published percentage. You need: a forecast that is getting less wrong, hours decided before the week rather than counted after it, overtime that always has a reason, sales per labor hour holding or rising, and turnover falling. Five directions, no targets.
What this course kept pointing at and deliberately did not cover is the one lever with the best return of all: people who are trained properly work faster, make fewer mistakes, need less supervision and stay longer — which touches every number in this course at once. How to hire, onboard and train so that happens is the next book in the catalog, and it is where the second batch of three finishes.