The fundamentals of shift scheduling — what every small-team manager learns the hard way

The first time I ever made a shift schedule for a small team, I made every classic mistake in about three days.

I scheduled someone for a shift that started thirty minutes after their previous one ended, in the same week. I forgot that the law in our jurisdiction required at least 11 hours of rest between shifts. I assumed that if someone clocked in at 11 p.m. and clocked out at 7 a.m., that was eight hours — which it is, but the payroll system disagreed because the shift “spanned two days” in the wrong way. I learned what “split shift premium” meant the day someone politely asked why their pay envelope was light. I had a schedule that looked clean on paper and produced two grievances in week two.

So this is the article I wish someone had handed me back then. It’s not exhaustive. It’s the five things I see new managers consistently get wrong, plus the practical workaround that fixes each one.

1. The shift that crosses midnight is one shift, not two

This is the single most common confusion in time tracking systems. Someone clocks in at 21:00 and clocks out at 05:00 the next day. That’s an 8-hour shift. Most software handles this correctly now — but if you’re using a spreadsheet, or an older system, or any tool that uses calendar dates as the primary key for shifts, you’ll find that the math breaks in subtle ways.

Why it matters:

• Weekly hour totals will be wrong if the shift is split across two day-totals.
• Overtime calculation, if it’s daily-rules-based, treats the shift as 3 hours on day 1 and 5 hours on day 2 — neither of which trips daily overtime.
• The overnight worker shows up twice in the day’s headcount.

The fix is to record shifts as atomic objects with a start timestamp and an end timestamp. Don’t store “shift on day X.” Store “started at 2026-04-26T21:00 ended at 2026-04-27T05:00.” Most modern shift-tracking apps, including FaceClock, do this by default. If you’re using a spreadsheet, structure it the same way and let the date columns be derived.

2. Split shifts are a category of their own

A split shift is when someone works, takes a long break (usually unpaid), then works again the same day. Restaurants do this constantly — a server works 11:00–14:30 for the lunch rush, leaves, comes back 17:00–22:00 for dinner. It’s two shifts in a sense. It’s one workday in another sense. In several jurisdictions it triggers a “split shift premium” — extra pay because the employee’s day was disrupted by a long unpaid gap.

The mistakes I see:

• Recording it as a single 11-hour shift with an unpaid break, which inflates the worked hours and may trigger overtime that isn’t actually there.
• Recording it as two separate shifts with no link, which makes weekly hour totals correct but loses the ability to detect the split-shift premium trigger.
• Recording it as one shift but having the time-tracking app auto-clock-out after a 90-minute idle, then the employee comes back, can’t recover, and a manager spends 20 minutes on Monday fixing it.

The right approach is to record two shifts but tag them as belonging to the same day’s split arrangement. If your system can’t do that, settle for two separate shifts and remember to apply the split-shift premium manually if your jurisdiction requires it. (California’s split-shift premium is the most common one US-side; several EU countries have similar rules under collective bargaining agreements.)

3. Overtime triggers vary more than people expect

The default mental model is “anything over 40 hours a week is overtime.” That’s the US federal standard. But it’s not the only standard, and even within the US, half a dozen states have different rules.

Some patterns worth knowing:

• Weekly overtime: anything over 40h/week (US federal default, many state defaults).
• Daily overtime: anything over 8h/day (California, some Canadian provinces).
• Daily and weekly stacked: California again — over 8h/day is overtime, over 12h/day is double-time, on top of the weekly rule.
• Bi-weekly schedules: some healthcare workers in the US can be paid based on an 80h/14-day rule rather than a weekly rule.
• Collective agreements: in much of Europe, the rules above are baseline, and the actual overtime threshold is set by industry-level collective agreements which can be substantially different.

What this means practically: don’t assume your time-tracking app’s overtime calculator is correct for your jurisdiction. Many of them have only the basic 40h/week rule wired up, sometimes with a configurable threshold. If you’re in California, in healthcare, or in Europe, do the overtime math in your payroll system rather than in the time clock. The time clock’s job is to give you accurate shift records. Calculation belongs downstream.

4. Rest periods between shifts are actually enforceable

Most jurisdictions have a minimum rest period between shifts. The EU Working Time Directive sets it at 11 hours minimum. Several Canadian provinces require 8. Some industries have higher minimums (truck drivers everywhere, for obvious reasons). The US federal level doesn’t impose this for most adults, but several states and many industry-specific regulations do.

If your scheduling tool doesn’t warn you when you’re scheduling someone for a sub-rest-period gap, you’ll occasionally do it without noticing. Especially around shift swaps. Especially around “can you stay late tonight and come in early tomorrow.”

The practical fix is a rule check at scheduling time. If you don’t have software that does that, keep a sticky note on the wall: “Minimum rest period: 11 hours.” Look at it every time you commit a schedule change.

5. Last-minute swaps will be 30% of your schedule, no matter how clean your initial draft is

Every new manager I talk to is surprised by how much of their scheduling time is spent on swaps after the schedule is published. Three days into the week, two people swap shifts. Five days in, someone calls in sick and you need a substitute. Six days in, a regular employee asks for a half-shift swap with a part-timer.

This is normal. It’s not a sign of bad scheduling. It’s a feature of any team that has more than four or five members.

What separates clean operations from messy ones:

• A clear single source of truth for the current schedule. Not a Google Doc, an email, and a paper sheet that disagree. One place. Everyone refers to it.
• A clear protocol for swaps. “Both people text the manager. Manager updates the schedule. Manager confirms back to both.” If swaps happen between employees without manager approval, you’ll inevitably hit a “but I thought you were covering!” situation.
• A way to detect and approve last-minute swaps in the time-tracking system. When the actual clock-in events don’t match the planned schedule, that’s data. You want to know that Maria covered for Sofia on Thursday because Sofia called in sick. Six months later, when Sofia’s manager asks why her shift count is low, the data is there.

Time tracking and scheduling are sometimes treated as the same software, sometimes as separate. Either approach works. What matters is that you can reconcile what was scheduled against what actually happened. A time-clock log that shows “Maria worked Thursday” is enough information, on its own, to do that reconciliation later.

The cost of a bad schedule, briefly

Bad scheduling shows up in places you don’t immediately associate with scheduling:

• Higher turnover. Employees who can’t predict their schedule a week ahead often leave.
• More errors during shifts (tired people, undertrained substitutes).
• Payroll disputes that eat manager time.
• Compliance fines or grievances when rest periods or overtime triggers are violated.
• Bad customer experience on understaffed shifts.

The cost of a good schedule is mostly time spent up front. Maybe an extra hour a week of attentive planning. The cost of a bad one is often invisible until it bites: a complaint from an experienced employee who quits the next month, or a payroll system that calculates overtime nobody noticed they were entitled to.

Where to start, if you’re new to this

If you’re managing your first team and the schedule is currently a piece of paper or a Google Doc that you wrote three weeks ago and have been quietly amending in red:

1. Write down the rules that apply to your jurisdiction. Maximum daily hours, minimum rest, overtime triggers, mandatory breaks. Print them out and put them on the wall.
2. Pick a single tool that holds the schedule. It can be free, but it has to be one place.
3. Pick a single tool that tracks actual clock-ins. (FaceClock fits this if you want it on-device; a simple sign-in sheet works too if you’re under five people.)
4. Set up one weekly hour to reconcile the schedule against the actual clocks. This catches errors early and gives you a feel for how your operation actually runs versus how you imagine it runs.

After a couple of weeks of doing this, the patterns become legible. You’ll know which employees show up on time. You’ll know which shifts always run long. You’ll know which days you’re consistently overstaffed. That’s where good scheduling starts — not from a template, but from data about your specific team.

That’s the boring truth: there’s no clever shift-scheduling shortcut. There’s just paying attention, consistently, for a while.