Blog

How Much Notice Is Fair Before Changing an Employee's Shift?

Treat a published schedule as a commitment, not a draft.

A fair default is to give at least as much notice as you'd want if the roles were reversed — and to aim for more than 24 hours whenever you can. Anything inside roughly a day is a favor you're asking of the person, not a change you're entitled to impose. And before you set any internal rule, check your local labor and "fair workweek" laws: in some places, predictability-pay rules set a legal minimum notice — and a penalty — that sits above whatever policy you write.

That's the short answer. The rest of this article makes it usable: how to decide what "fair" means for a specific change, why swaps, cancellations, and true emergencies carry different obligations, and what to do when you genuinely can't give enough notice.

The core principle: a published shift is a promise

The moment you publish a schedule, your team plans their lives around it. They arrange childcare, book the dentist for their day off, turn down a second job's hours, tell family when they're free. A published shift isn't a suggestion you can quietly revise — it's a commitment they've already acted on.

That reframing does most of the work. Once you treat the published schedule as a promise, the fairness question stops being "what am I allowed to do?" and becomes "what am I asking this person to give up, and how much warning does that deserve?"

Two rules of thumb that hold up in practice:

• The reciprocity test. Would you accept this notice window if your manager did it to you? If you'd be annoyed to lose a shift with three hours' warning, your team will be too.
• The published-vs-draft line. Before a schedule is published, move things freely — that's what drafting is for. After it's published, every change costs someone something. Publish later-stage schedules only when you're confident in them, so you're not "editing" plans people have already built their week around.

A decision tree: what kind of change is this?

Not every schedule change carries the same obligation. Before you message anyone, figure out which of these three you're actually doing — because the fair response is different for each.

1. A swap (the employee wants the change)

If the person on the shift is the one who wants out — they found coverage, they have a conflict, they'd prefer a different day — the fairness burden is low. They're choosing this. Your job is mostly to make it easy and to make sure the swap is qualified: right role, right certifications, no one pushed into overtime or an illegal turnaround (a "clopen" — closing then opening a few hours later).

Fair practice:

• Let staff initiate and arrange swaps themselves, then approve rather than broker every one by hand.
• Confirm the replacement is actually qualified and available before you sign off.
• Keep a record of who's now responsible, so no one shows up to a shift they thought they'd handed off.

Swaps are the cheapest kind of change because consent is built in. The more of your changes you can convert into swaps — by giving staff a self-serve path — the fewer changes you have to impose.

2. A cancellation or reduction (you're taking hours away)

This is where fairness gets serious, because you're removing income the person was counting on. Sending someone home early or cutting a shift because it's slow is one of the most resented moves in shift work — and in some jurisdictions it's regulated.

Fair practice:

• Give the most notice you possibly can. Cutting a Saturday shift on Thursday is a disappointment; cutting it two hours before is a broken promise.
• Offer, don't order. "It's looking slow — anyone want to drop tonight's shift?" turns a loss into a choice. Some people will happily take the night off; let them self-select before you cut anyone involuntarily.
• Know your predictability-pay exposure. Several "fair workweek" laws require you to pay a portion of a cancelled or shortened shift if you didn't give enough notice. If your area has one, that penalty is the floor — don't design a policy under it.

3. A coverage emergency (someone must come in, or someone must go)

A no-show, a sudden illness, a burst pipe, a double-booking. Here you genuinely need to change the plan on short notice — and everyone knows emergencies happen. The fairness question shifts from "how much notice" to "how you ask."

Fair practice:

• Ask, frame it as the favor it is, and remember it. "I'm in a bind — any chance you can come in tonight? I know it's last minute." Then actually track who bailed you out and return the goodwill.
• Never disguise a planning failure as an emergency. If you're calling people in at the last minute every week, that's not an emergency — it's an under-built schedule. Staff can tell the difference, and the "emergency" tone wears out fast.
• Have a standing volunteer list. People who want extra hours are your first call. Reaching a willing person beats pressuring an unwilling one every time.

What to do when you can't give fair notice

Sometimes you simply can't. Someone quit, someone's in the hospital, and you need a body tonight. Fairness doesn't mean never asking — it means asking honestly and squaring up afterward.

1. Name it as a favor, not a directive. The tone tells the person whether you respect their time. "Can you help me out?" lands completely differently than "You're on at 5."
2. Make the ask opt-in first. Broadcast to people who want hours before you lean on anyone specific. Willing coverage is faster and fairer.
3. Give something back. First pick of next week's shifts, the shift they wanted, a genuine thank-you in front of the team. Last-minute goodwill is a loan; repay it.
4. Then fix the upstream cause. One real emergency is life. A pattern of them is a scheduling problem — publish earlier, staff a small buffer, or build a proper on-call list.

How far ahead should you publish in the first place?

Most last-minute changes trace back to publishing late. The earlier a schedule goes out, the more runway everyone has to flag conflicts before it's a promise you have to break — and the fewer imposed changes you make all week.

Aim to publish at least one to two weeks out. Where fair-workweek laws apply, that advance-notice window may be legally required (often around two weeks), with penalty pay attached if you post late or change a posted shift. A useful sequence:

1. Draft early and let staff submit availability and time-off before you finalize.
2. Publish with real lead time so conflicts surface while the schedule is still editable.
3. Give staff a self-serve way to swap, so most post-publish changes are their choice, not your imposition.

Do those three things and the hard question in this article — "how much notice before I change someone's shift?" — comes up far less, because you're changing far fewer shifts against people's will.

Where a scheduling tool helps

You can run all of this on a spreadsheet and good habits. But the friction — collecting availability, publishing on time, brokering swaps by hand, tracking who covered what — is exactly what pushes managers into last-minute changes.

Zentru is a scheduling and operations platform for shift-based teams (restaurants, bars, hotels, gyms, cruise ships, hospitals, and other multi-location businesses). It's built to shrink how often you impose changes at all: publish schedules earlier, collect time-off and availability up front, and give staff a self-serve shift-swap path so most changes are their choice, not your emergency. Its "Ask Amil" feature lets you describe next week in plain language and get back a validated draft schedule you approve or edit — instead of dragging shifts by hand, which makes publishing early realistic instead of a good intention. Zentru handles scheduling, time-clock, and hours export (so you can hand clean hours to your payroll provider), but it doesn't run payroll or set your local labor rules — treat it as the tool that makes fair notice easier to give, not a substitute for knowing your obligations.