You published a schedule. Something else happened. The distance between those two things is where the money and most of the frustration live, and unlike almost everything else in a restaurant, it is easy to measure — the tool is already recording it.
📊 The four numbers
| Number | What it tells you | What to do about it |
| Scheduled hours vs paid hours | Whether your plan survives the week. A consistent overrun is early clock-ins, late clock-outs or shifts you added on the day | Find where it happens. It is usually one station and one hour |
| Punch edits | How often reality had to be corrected by hand | A few is normal. A lot means the clock is inconvenient, or something else |
| Late arrivals and no-shows | Who, and when. Almost always concentrated in a few people and a few shifts | A conversation about a pattern, not about last Tuesday |
| Swap and cover requests | Which shifts nobody wants, and who is carrying them | Fix the shift, or accept that it is a shift you have to pay more attention to |
Fifteen minutes a week, at the same time you approve hours. This is not analysis; it is looking at four numbers and asking whether anything moved.
🔍 Reading the swap report
The one people ignore, and the most diagnostic of the four. A shift that generates constant swap requests is telling you something specific:
- Always the same shift? There is something wrong with it — the section, the hours, who runs it, or the closing duties that follow.
- Always the same person giving away shifts? Their availability has changed and they have not said so. Ask; it is usually a class or a second job, and it is fixable on paper.
- Always the same person taking them? Someone wants more hours than you are giving them. That is useful — they are your standby list and possibly your next promotion.
- Swaps rising overall? The schedule is drifting away from people's real availability. Time to re-collect it.
🔁 The weekly and quarterly routine
Weekly, fifteen minutes:
- Approve hours and look at the four numbers.
- Read the swaps and note anything repeating.
- Build next week from the template, not from memory.
- Publish on the same day, at the same time.
Quarterly, an hour:
- Re-collect availability. It has changed and nobody told you. This one hour prevents a season of swaps.
- Compare your staffing template against the real sales curve. Peaks move. A template built in spring is wrong by August.
- Audit shift distribution. Who has been getting the good shifts, from topic 5.
- Check the tool is still being used as designed. If a paper copy has reappeared on the office wall, find out why — it is telling you something the app is not doing.
🎯 What good looks like, and what comes next
You do not need a target number for any of this. You need direction: fewer messages, fewer surprises, a smaller gap between planned and paid, and staff who know their hours two weeks out. If those four are moving the right way, the schedule is working, whatever the software costs.
What this course deliberately did not do is put a price on any of it. Hours became money the moment they were worked, and how to read that — labor cost as a percentage, forecasting hours against sales, what overtime really costs, what turnover costs you every time somebody leaves — is the next book in the catalog. It picks up exactly where this one stops, with the same reports you have just learned to read.