How to Schedule a Small Restaurant Team (Under 10)

Labour is climbing toward 31-32% of revenue for Canadian restaurants in 2026. For a small independent doing $40,000 a month, that's $12,400 walking out the door in wages, payroll taxes, and benefits before you've paid rent or ordered a single case of produce.

But here's what most scheduling advice gets wrong: it assumes you have 30 staff across three dayparts and a manager whose only job is building the schedule. You have eight people. Maybe six. Maybe it's you, your partner, and four employees you're terrified of losing.

This guide is for that restaurant. No software pitch. No enterprise frameworks. Just the math and the methods that actually work when your team fits around one table.

Why scheduling is a labour cost problem, not a time management problem

Most operators think of scheduling as logistics: who works when. That framing misses the money.

A 40-seat restaurant running two services a day with an 8-person team carries roughly $12,000-$13,000 in monthly labour costs at current Canadian wages. The difference between a good schedule and a sloppy one is typically 6-10% of that spend. That's $720-$1,300 a month you're either keeping or burning on overlap, overstaffing slow shifts, and overtime you didn't see coming.

Three places the money leaks:

Shift overlap. Ten extra minutes of overlap between the lunch closer and the dinner opener, five days a week, across three employees. That's 13 hours a month of double coverage you didn't need. At $18/hour, that's $234 gone.

Habit-based scheduling. You scheduled four people for Tuesday lunch because you've always scheduled four people for Tuesday lunch. But your POS data (or your memory, if you're honest with yourself) says Tuesday lunch averages 22 covers. Three people can handle that.

Untracked overtime. Ontario's overtime kicks in at 44 hours per week. If one employee is at 42 hours by Thursday evening and you ask them to cover Friday morning, you're paying time-and-a-half for a shift that a cross-trained teammate could have handled at regular rate.

How to build a demand-based schedule without software

You don't need a $200/month platform to schedule based on demand. You need last month's numbers and a spreadsheet.

Step 1: Pull your covers by daypart. Most POS systems export daily transaction counts. If yours doesn't, count manually for two weeks. You need average covers for each service window: weekday lunch, weekday dinner, weekend lunch, weekend dinner, plus any brunch or late-night you run.

Step 2: Set your covers-per-staff ratio. For full-service, 15-20 covers per front-of-house staff member is a common range. For the kitchen, it depends on your menu complexity, but a rough starting point is 25-30 covers per cook. These aren't universal rules. Adjust based on your menu, your layout, and how fast your team actually moves.

Step 3: Map staff count to each shift. A weekday lunch averaging 35 covers needs two FOH and two BOH. A Saturday dinner averaging 80 covers needs four FOH and three BOH. Write it down. This is your baseline.

Step 4: Build in flex. One person on your team should be the designated flex position for each shift. They come in if covers are tracking above average by 11:30 a.m. (for lunch) or 5:30 p.m. (for dinner). Text them at the trigger point. This is where cross-training matters most: your flex person needs to work either side of the pass.

Shift	Avg. Covers	FOH	BOH	Flex?	Notes
Mon-Thu Lunch	30-35	2	2	No	Owner covers host
Mon-Thu Dinner	50-60	3	2	Yes	Flex FOH on call
Fri Dinner	75-85	4	3	No	Full team
Sat Dinner	80-90	4	3	No	Full team
Sun Brunch	60-70	3	2	Yes	Flex BOH on call

Cross-training as a scheduling strategy

Cross-training usually gets pitched as a development tool. For small teams, it's a scheduling tool. Every person who can only work one station is a scheduling constraint. Every person who can work two stations is scheduling freedom.

53% of Canadian restaurant operators cross-trained BOH and FOH staff in 2025. That number is climbing because operators figured out that a server who can expo during a rush is worth more than a server and an expeditor scheduled separately.

The skill matrix approach. Build a simple grid. Employees down the left. Stations across the top: host, serve, bar, expo, prep, line, close. Mark each person's proficiency: primary (P), secondary (S), or learning (L). Your scheduling constraint is visible in one glance.

Employee	Host	Serve	Bar	Expo	Prep	Line	Close
Alex	P	P	S	S	·	·	P
Jordan	S	P	P	·	·	·	S
Sam	·	·	·	P	P	P	S
Taylor	·	S	·	S	P	S	P
Morgan	P	S	·	P	S	·	S

When you look at this grid and see a column with only one "P," that's your vulnerability. If Sam is your only line cook and Sam calls in sick on Saturday, you're in trouble. The fix isn't hiring. It's getting Taylor from "S" to "P" on line over the next month.

How to actually cross-train without disrupting service. The worst time to train is during a rush. The best time is the 20-30 minutes before service when the kitchen is prepped and the floor isn't moving yet. A server shadowing the line cook during prep learns plating, timing, and menu flow without slowing anyone down. Fifteen minutes before lunch service is a window most operators waste. Use it.

Pair it with slow shifts. Tuesday lunch with 22 covers is when your FOH person learns to expo. Not Friday night with 85 covers. Schedule the training shift, don't just hope it happens.

Incentivize it. A $0.50-$1.00/hour bump for each secondary station mastered costs you $80-$160 a month per employee. Compare that to the cost of calling in an agency cook at $25-$30/hour when your only line cook is sick. Cross-training pays for itself the first time it saves you from a staffing emergency.

Split-shift math: when it works, when it doesn't

Split shifts let you staff the lunch rush and the dinner rush without paying someone to stand around from 2:00 to 5:00 p.m. For a small team, that dead zone is pure cost. But split shifts come with provincial rules you need to know.

The basic math. An employee working 11:00 a.m. to 2:00 p.m. and 5:00 p.m. to 10:00 p.m. gives you 8 hours of productive coverage across both peaks, with zero hours of paying someone to polish glasses during the gap. Compare that to a straight 10:00 a.m. to 6:00 p.m. shift that covers the lunch rush and the pre-dinner lull but misses the dinner peak entirely.

Provincial rules at a glance:

Province	Key Rule	Watch Out For
Ontario	Three-hour rule: if employee reports and works less than 3 hours, pay for 3 hours	Each segment of a split shift must meet the 3-hour minimum if the employee regularly works 3+ hours daily
BC	Split shift must complete within 12 hours; minimum 2 hours per shift segment	The 12-hour window limits how far apart you can space the segments
Alberta	If break between segments exceeds 1 hour, minimum compensation applies to each segment	The per-segment minimum means very short segments cost more than they're worth
Quebec	CNESST: employee must have 8 consecutive free hours between two shifts	Limits back-to-back split shifts on consecutive days

When split shifts make sense: You have a true dead zone (sub-10 covers between 2:00 and 5:00 p.m.), employees live close enough that the commute home and back isn't punishing, and you're running both lunch and dinner service.

When they don't: Your team has long commutes (they'll resent the unpaid gap), you only run dinner service (no need to split), or you're in a province where the per-segment minimums eat into the savings.

One more thing: split shifts burn goodwill faster than almost any other scheduling practice. If you're using them, make sure the employees on splits are choosing them, not being forced. Some people love the midday break. Others will quietly start looking for a new job.

The weekly template: stop building from scratch

If you're building a new schedule from zero every week, you're wasting 2-3 hours and probably making worse decisions than a template would.

Build a base template for your standard week. Monday through Thursday looks roughly the same. Friday and Saturday are your peak template. Sunday is its own thing. The template covers 80% of your weeks. You adjust for holidays, events, and staff requests, not rebuild from nothing.

What a template looks like for an 8-person team running lunch and dinner:

The template lives in a shared Google Sheet or printed on a clipboard in the kitchen. Everyone can see it. Requests for changes go in by Wednesday for the following week. You finalize and post by Thursday evening, giving everyone at least 10 days of notice. Publishing 10-14 days in advance reduces last-minute swaps by roughly half.

Track hours weekly, not monthly. By the time you catch overtime on a monthly review, you've already paid it. A simple column in your spreadsheet that totals each person's hours by Wednesday afternoon lets you redistribute Thursday and Friday shifts before anyone hits the overtime threshold. For Ontario, that threshold is 44 hours per week. For most other provinces, it's 40.

Five scheduling mistakes small teams make

1. The owner absorbs everything. You're already working 60 hours a week. Taking extra shifts because "it's easier than finding coverage" is a short path to burnout and a longer path to closing. Schedule yourself like an employee. Set a hard cap. Stick to it.

2. Ignoring the close. Closing shifts are the least desirable and the hardest to staff. If the same two people always close, they'll leave. Rotate closing across the team. Use the skill matrix to make sure everyone has a "C" in the close column.

3. Scheduling by seniority instead of demand. Your best server gets the Friday night shift because they've been there longest. But your second-best server is nearly as good and costs $2/hour less. On a 70-cover Friday, the difference in service quality is negligible. On your monthly labour cost, it's not.

4. No backup plan. A small team with no cross-training and no on-call flex position is one sick call away from chaos. The skill matrix and flex system above exist specifically for this. Build the plan before you need it.

5. Clopening. Scheduling someone to close at 11:00 p.m. and open at 7:00 a.m. the next day. Besides being terrible for the employee, it produces worse work. A tired prep cook makes mistakes that cost you in food waste. Most provinces require 8 consecutive hours off between shifts. Follow the rule, and follow common sense.

What about scheduling software?

For a team under 10, you probably don't need it yet. A Google Sheet template, a group text thread for swaps, and a printed schedule on the kitchen wall cover the basics.

If you're spending more than 2-3 hours a week on scheduling, or if shift swap requests are becoming unmanageable, that's when scheduling software starts earning its cost. Sling offers a free tier. Agendrix is Canadian-built and starts at around $3/user/month. 7shifts is restaurant-specific and popular with Canadian operators.

But don't buy software to solve a process problem. If your scheduling is broken because you don't know your covers-per-shift or haven't cross-trained anyone, software will just automate the same bad decisions faster.

The bottom line

Labour costs are projected at 31-32% of revenue for Canadian restaurants in 2026. You can't control minimum wage increases. You can't control the labour shortage. But you can control how you deploy the team you have.

Demand-based scheduling, cross-training, and a simple weekly template won't turn a struggling restaurant around on their own. But they're the difference between running at 32% labour cost and running at 29%. On $40,000 a month in revenue, that's $1,200 back in your pocket. Every month.

Start with the skill matrix. Build it this week. You'll see your scheduling constraints, your training priorities, and your coverage gaps in one glance. Everything else follows from there.

When you're ready to take reservations, Trudy's Table is built for Canadian independents.

Sources: Snappy: 2026 Labour Cost Projections, Restaurant365: Cross-Training Staff, Ontario ESA: Three-Hour Rule, BC Employment Standards: Hours of Work, Alberta Employment Standards: Hours of Work, Sling: Restaurant Scheduling Tips, Escoffier: Training Without Disrupting Service.

---