What should a shift app check before a schedule is published?

At minimum: nobody is double booked across two shifts, nobody is scheduled during declared unavailability or approved leave, every shift has the roles and certifications it requires, minimum rest between consecutive shifts is respected, and the projected hours per person are visible before the schedule goes out rather than after payroll runs.

How should shift swaps work in a scheduling app?

A swap needs three states, not one: offered, accepted by a colleague, and approved by whoever owns the rota. Letting two staff agree between themselves without the third state is how a shift ends up uncovered, because the schedule of record never changed. The app should also re-run the same conflict checks on the person picking the shift up.

Does a work shift app need to work offline?

Reading does, editing usually does not. Staff check their next shift in places with poor signal, so the current rota should be readable from a cached copy. Publishing and approving are manager actions taken on a laptop or in an office, so they can reasonably require a connection.

Do predictive scheduling laws affect what the app has to record?

In several United States jurisdictions, yes. Oregon's Fair Work Week law requires covered employers to give written work schedules in advance and to keep those records, and San Francisco's Formula Retail Employee Rights Ordinances impose advance notice and predictability pay. If you operate somewhere with such a rule, the app has to keep a timestamped history of what was published and when it changed, because that history is the evidence.

Can I build my own work shift app instead of buying one?

It is a reasonable option when your rota rules do not fit an off-the-shelf product, which is common for teams with unusual role requirements, split shifts, or approval chains that cross departments. Describing the rules you actually run and generating an app around them avoids bending your process to fit someone else's schema.

Articles · App ExamplesUpdated September 2026

A work shift app is a rulebook, not a calendar.

Any work shift app can show who is on at two o'clock. The ones worth having refuse to publish a schedule that breaks a rule: the same person in two places, a shift with nobody qualified to close, a closing shift followed by an opening nine hours later. That difference is what separates a shared calendar from real employee management apps.

This page is about the rules underneath the rota: what the app has to check before you hit publish, what a swap actually has to do to the schedule of record, and why the history of changes matters more than the grid on screen.

See the rules a schedule has to enforce

The short version

Six checks before publish, and a history nobody can quietly edit.

A work shift app earns its place by refusing bad schedules. It should block a double booking, respect declared unavailability, insist that every shift has the roles it needs, hold a minimum rest gap, show projected hours before payroll finds them, and apply hour limits for staff under 18 without being asked.

Then it has to remember. Every publish, every swap, every last-minute change, with a timestamp and a name against it. In some jurisdictions that record is not an internal nicety, it is the evidence you would need if someone asked.

What the app has to check before you hit publish

Every rule below exists because a real shift went uncovered. The useful question is not whether your manager would spot the problem on a careful read. It is whether the schedule can be published at all while the problem exists.

The two that are non-negotiable

Double booking and declared unavailability. Both are absolute: there is no version of the schedule where either is acceptable, so the app should refuse the save rather than warn and let it through. A warning that can be dismissed is a warning that will be dismissed at 11pm on a Friday.

The one everybody forgets

Role and certification coverage. A shift can look fully staffed and still be broken, because the person who can open the safe, sign off a temperature log, or supervise a minor is not on it. This is the rule that turns a headcount into a schedule, and it is the one spreadsheets never enforce.

Try it

Which of these does your schedule enforce today?

Turn on the rules your current rota actually checks by itself, without somebody noticing. Not the ones a good manager catches by reading it twice.

0 of 15 points · Right now the schedule is whatever someone typed. Every rule below is a mistake it cannot catch.

A swap is three states, not an agreement

The most common failure in shift scheduling is not a bad rota. It is two people agreeing a swap between themselves while the schedule of record never changes. On the day, the rota still says one name and the other person is at home. Nobody lied. The system simply had no idea.

The fix is to model a swap as three distinct states: offered, accepted by a colleague, and approved by whoever owns the rota. Only the third state changes the schedule. The same conflict rules then have to run again against the person picking the shift up, because their availability is exactly what nobody checked. That flow is worth designing on its own, which is why a shift swap app is often the first thing teams build once the base rota works.

Open shifts follow the same shape. Posting an uncovered shift to everyone and taking the first acceptance is fine, as long as acceptance is a claim that still has to pass the checks, not an instant booking.

The schedule is not the timesheet

A rota says what was meant to happen. Payroll needs what did. Keeping those two things in one field is how a business ends up paying scheduled hours instead of worked hours, or arguing about a late finish nobody recorded.

Under the Fair Labor Standards Act, United States employers have to keep records for each non-exempt worker including the hours worked each day and the total hours each workweek. The Department of Labor is explicit that no particular form is required, but the information has to be accurate. A schedule alone does not satisfy that, which is why the clock is a separate record and often a separate time clock app for employees.

US Department of Labor, Fact Sheet 21 on FLSA recordkeeping

Where the schedule itself becomes a legal record

Predictive scheduling rules change what your app has to store. In Oregon, employers with at least 500 employees worldwide in retail, hospitality or food services must give workers a written work schedule at least 14 calendar days in advance, posted and easily visible, covering all work and on-call shifts. If the employer then changes it without notice, additional compensation is owed: one hour at the regular rate for adding more than 30 minutes to a shift or adding a shift, and one half the regular rate per scheduled hour for hours taken away or a cancelled shift.

Read that as a data requirement rather than a legal one and it is clear what the app owes you: the exact schedule as published, the timestamp it was published at, and every change since with who made it and when. Without that history there is no way to work out what is owed, or to show that nothing is.

Similar ordinances exist in other United States cities and states, and the thresholds and industries differ in each. Check the rule where you actually operate rather than assuming Oregon's numbers apply.

Oregon Bureau of Labor and Industries, predictive scheduling

What each option actually gives you

Option	Blocks conflicts	Keeps a change history	Swaps	Fits your rules
Printed rota on the wall	No	No	by conversation	yes, it is whatever you write
Spreadsheet	No	only if nobody overwrites it	by conversation	yes, until the tabs multiply
Group chat	No	buried in messages	informal, often unrecorded	no schedule of record at all
Off-the-shelf scheduling platform	Yes	Yes	built in	their rules, not yours
A shift app you build	Yes	Yes	the approval chain you actually use	yes, by definition

Building one around your own rota

Most teams do not leave a scheduling platform because it lacks features. They leave because their rules are not in it: a split shift that counts as one, a certification that only three people hold, an approval that has to go through two departments. The workaround is a spreadsheet kept alongside the tool, which is the clearest sign the tool does not fit.

Newly is an AI app builder: you describe the app you want, including the rules above in plain English, and it builds and ships a real mobile app you own. Plans start at $25 a month, there is no free plan, and iOS builds ship through App Store Connect. It is not a scheduling product and does not compete feature for feature with one. It is what you use when the product you tried does not match the way your rota actually works.

One thing to decide early rather than late: how the team hears that a schedule changed. A published rota nobody opens is not a published rota, so the delivery mechanism matters as much as the grid, and it is worth reading up on sending the shift reminder before you design the publish button.

Questions people ask about shift apps

A work shift app is the tool a team uses to build a rota, publish it, and change it afterwards. It holds who is working when, which roles each shift needs filled, who is unavailable, and a record of every swap and cover, so the published schedule and the hours actually worked can be reconciled later.

Describe your rota, rules and all

Write down how your schedule actually works, including the rule everybody has to remember, and build the app around it.

Start building