Every business with a rota and a clock has two numbers for the same week: the hours it scheduled and the hours it actually paid. Most look at the second one, on the payroll run, once a month. The interesting information is in the difference, and it decays fast - a variance you spot on Monday is a conversation, the same variance spotted six weeks later is an argument about what happened.
Read the sign first
Actual consistently higher than scheduled is the common case. Shifts overrun. Read it as one of four things:
- The workload genuinely exceeds the plan and your coverage requirement is wrong.
- Handover isn't built into the shift, so every shift ends late by ten minutes.
- One or two individuals habitually stay, either because they want the hours or because they can't say no.
- Closing procedures take longer than the schedule assumes - often because they were timed when the business was smaller.
Each has a different fix, and the fix is almost never "tell people to leave on time."
Actual consistently lower than scheduled looks like good news and usually isn't. Either the shift is genuinely overstaffed - fine, adjust the plan and take the saving - or people are clocking out and then finishing up, which is unpaid work and a liability.
Actual roughly equal to scheduled, every week, to the minute is the pattern to be most suspicious of. Real work has variance. A perfectly flat variance usually means the clock is being filled in from the rota rather than from reality, which means your time record is a copy of your schedule and proves nothing.
Cut the number four ways
A single site-wide variance figure hides everything. The same total can come from very different places, so slice it:
By person. One name accounting for most of the overrun is a workload or boundary conversation, not a scheduling change. Look particularly for the person whose actual hours exceed scheduled every single week - that is the "always says yes" pattern showing up in payroll.
By day of week. Overrun concentrated on Fridays and Saturdays means your peak coverage is wrong. Overrun spread evenly means your shift lengths are wrong.
By shift type. Closing shifts overrunning while opening shifts run to time is a closing-procedure problem with a specific, fixable cause.
By start time. If every shift starts eight minutes early, you have a queue at the terminal or a culture where arriving on time counts as late.
Turn it into money
Variance in hours is abstract. Multiply it by the relevant rate and it stops being abstract.
Twelve people averaging fifteen minutes of unplanned overrun per shift, five shifts a week, is fifteen hours a week. At €15 an hour that is roughly €11,500 a year - enough to fund a meaningful part of another person, spent instead on the least efficient hour of everybody's day.
Do this calculation once and the weekly review stops feeling like admin.
What to do with what you find
The temptation is to police the variance. Resist it. Variance is a measurement of the difference between your plan and reality, and when they disagree, reality is not usually the thing that is wrong.
The productive sequence is:
- Find the pattern. One person, one day, one shift type, one site.
- Ask what the shift is actually doing in the overrun window. Nine times out of ten it is a real task the schedule doesn't account for.
- Change the plan to include it, or change the process so it isn't needed.
- Re-measure next week.
If after that the variance persists in one individual and not in the same shift worked by others, then it is a personal conversation. But that is the fourth step, not the first.
Make it a ten-minute habit
Once a week, same slot, before you build the next rota:
- Total scheduled hours, total actual hours, variance in hours and in money.
- The three largest individual variances.
- Anything that changed from last week.
Ten minutes. It is the highest-leverage ten minutes in the operational week, because it is the only routine that connects the plan you make to the money you spend.