Your Staff Schedule Is a Labour Cost Decision

Most schedules get written the same way every week: open last week's, change the names that asked for a day off, post it. Nobody decided anything. But that piece of paper is the single biggest controllable cost in the business, and writing it by habit means your labour percentage is being set by inertia.

Food cost gets scrutinized to the penny. Owners argue with suppliers over a case of tomatoes. Then they schedule four servers on a Tuesday because that's what they did last Tuesday, and nobody checks whether Tuesday needed four.

This is not about cutting hours until service breaks. It's about putting the hours where the sales are.

Pull your sales by hour, by day

Every POS worth using can produce this. You want a grid: days of the week across the top, hours down the side, sales in each box. Four to six weeks of data, averaged. Skip any week with a snowstorm, a holiday, or a street festival — those distort the picture.

Print it. Put it next to the schedule.

What you're looking for is the shape of your week, not the total. Two restaurants can both do the same weekly sales, one with a flat spread and one with a spike Friday and Saturday night and near-nothing Monday. Those two businesses need completely different schedules, and if you've never mapped it you're probably staffing the average instead of the reality.

Then do the same for covers or transaction counts, not just dollars. A Sunday brunch doing decent sales on large tables is a different labour problem than the same dollars spread across forty solo diners. Dollars tell you revenue. Covers tell you work.

Work out what an hour of labour has to earn

Here's the number that changes how you read the grid: sales per labour hour.

Take the sales in a given hour. Divide by the number of staff hours you're paying for in that hour — everyone, front and back, including the person doing prep in the basement. That's what each hour of labour produced.

Do it for your busiest Saturday hour and your deadest Tuesday hour. The gap will be bigger than you expect. That gap is where your money goes.

You don't need a target borrowed from an industry guide. You need your own baseline: what does an average hour produce across the whole week, and which hours fall so far below it that they're being paid for by the good hours? Once you can see that, the schedule stops being a calendar and starts being a set of decisions.

Staff to the shape, not the shift

The habit that costs the most is the full shift. Everyone comes in at four, everyone leaves at close. Meanwhile your sales grid says the room is quiet until six and dead after nine on a Wednesday.

Things that actually work:

Staggered starts. One server at open, second at the point your grid shows the climb starting, third only on the days that justify a third. Kitchen the same way. The cook who doesn't need to be there until five shouldn't be clocked in at three-thirty out of tradition.

Cut order set in advance. Decide before service who leaves first if the room is slow, and tell them. Staff hate being cut on a whim. They tolerate it fine when they knew going in that they were the early cut and the closer gets the late money.

One person covering two stations in the slow hours. The host who runs the takeout counter until seven. The prep cook who also handles the first orders of the night. This only works if you've trained for it, which means training is a scheduling tool.

Move prep off the peak. Prep hours are the most schedulable hours you have because they aren't tied to customers. If your prep person can come in Tuesday morning instead of Friday afternoon, you've moved labour out of an expensive hour and into a cheap one at no cost to service.

The shoulder hours nobody wants

Two to five in the afternoon. Nine-thirty to close on a weeknight. The hours that are too slow to need a full team and too busy to close.

Your options are honest ones:

1. Close. If three to five produces almost nothing, look at what you lose against what you save. Some rooms genuinely should not be open. Check your Google Business Profile hours match whatever you decide — mismatched hours cost you more than the shift does.
2. Run it on a skeleton. One person, limited menu, and a clear policy about what happens if six people walk in. Make sure whoever's alone has a way to call for backup.
3. Give it something to do. Shoulder hours are where the phone gets answered, the catering emails get returned, the online orders get packed. If the labour is already there, the question is whether it's producing anything besides waiting.

That third option is where a lot of owners find the real win. The two-to-five person who handles the phone and the pickup orders isn't dead labour — they're the reason you're not losing calls at five-thirty. We wrote more about how that plays out in what to do with the phone during your dinner rush. If the phone and the online orders can largely run themselves, the shoulder shift gets a lot cheaper to justify — that's most of what our restaurant setup is built to handle.

Ontario rules that catch owners out

These are the ones that turn a scheduling decision into a payroll problem. Treat this as a prompt to check, not as legal advice — confirm the current details with the Ontario Ministry of Labour or the Employment Standards Act guide before you rely on any of it.

The three-hour rule. If you schedule someone who normally works more than three hours, and they show up ready to work but you send them home early, you generally owe them three hours' pay. This is the one that punishes optimistic scheduling. If you schedule five people hoping it's busy and cut two at the door, you paid for those hours anyway. Better to schedule four and have someone on call who gets paid only if they come in.

Rest between shifts. There are minimum rest requirements — time off between shifts, and time off in each work week or two-week period. The classic violation is the close-then-open: someone finishes at one in the morning and is back for a ten o'clock open. It's not just hard on them, it may not be allowed.

Eating periods. An employee generally gets an uninterrupted eating period after a stretch of consecutive hours. "Grab something when you can" isn't a break.

Public holiday pay. Ontario has nine public holidays. Most employees qualify for public holiday pay, and what they get depends on whether they worked, whether they agreed in writing to work, and their earnings in the weeks before. Those days are also usually your busiest — so a holiday is both a revenue day and a labour-cost day, and it should be scheduled deliberately.

Keep the records. Hours worked, wages paid, the written agreements. If there's ever a complaint, the records are the whole case.

Make it a weekly habit

Build the schedule against the sales grid, not last week's schedule. Then, at week's end, compare what you planned to what actually happened — sales by day against hours by day. Where was the mismatch?

That comparison takes fifteen minutes and it's the part almost nobody does. Without it, you never learn whether Wednesday really needed the extra cook.

One more thing: tell your staff what the shape of the week looks like. Not your cost numbers — the pattern. People who understand why Tuesday is a two-person night stop reading it as favouritism. And the server who knows Thursday is climbing will tell you before your POS does.

The schedule isn't the calendar. It's the budget.