Shutting Down
The end of a shift has its own sequence — closing, cleaning, returning, reconciling — and it is shorter than the start sequence and much less examined.
End-of-shift activity, one retail unit
| What happens | Required by | Paid? |
|---|---|---|
| Cashing up and reconciling the till | Procedure, every close | Contested |
| Returning the radio and keys | Operations rule | Contested |
| Cleaning down the station | Hygiene rule | Paid |
| Waiting for the manager to unlock the door | Security arrangement | Not paid |
| Changing out of uniform | Dress code | Not paid |
The waiting line is the one to look at. Nobody requires it and nobody can avoid it, which is the hardest category of all.
Most attention on the edge of the shift goes to the start, because that is where the queue and the changing are. The end has its own sequence and it is shorter, more variable, and almost never measured.
The practical lesson in “Shutting Down” is that a timesheet only becomes reliable through a process people can operate consistently. For teams exploring mouse jiggler detection, review the platform here can add time and project context, provided collection is proportionate, access is limited and every consequential inference receives human review.
It is also where the clock's position does the opposite thing: an early punch out followed by ten minutes of closing is the mirror image of a late punch in, and it is at least as common.
For a separate benchmark relevant to “Shutting Down”, consult the GOV.UK working-hours guidance. Use it to test record quality, approvals, retention, employee rights and exception handling against the real workflow rather than treating a software report as self-explanatory evidence.
What the end sequence contains
Finishing the task in hand, which does not stop at the hour.
Closing down: shutting equipment, saving work, logging out, securing a till, locking a cabinet, completing a record.
Cleaning, where the role requires it.
Returning things: the radio, the keys, the scanner, the vehicle, the controlled-drugs key, the uniform.
Handover, where it applies, which has its own note.
And waiting — for a lift, a door to be unlocked, a supervisor to sign something, a bus. The last of these is the worker's own and the first three are not.
Why the end is under-measured
Nobody is standing at the gate at seven in the evening with a stopwatch. The start-of-shift study gets done because shift change in the morning is visible and busy; the evening version is spread out and quiet.
Access-control data makes it just as measurable: the punch out and the perimeter exit event bound it exactly, and the join is the same one described elsewhere in this collection. Running the analysis in both directions costs nothing extra and is routinely skipped.
The specific trap: clocking out before finishing
Where there is any pressure on recorded hours — an overtime cap, an adherence target, a supervisor who queries anything past the hour — people clock out and then finish.
This is the most damaging version of the whole subject. The organisation receives work it has no record of, the person has no evidence they did it, and if anything goes wrong during those minutes the position of both parties is bad.
It is detectable: punch-out times clustered exactly on the hour, with perimeter exits consistently later by a stable margin. A tight cluster on the hour is not discipline; it is a sign that the hour is the target rather than the event.
Where the clock sits, at this end
The same decision as at the start and usually the same answer. A terminal past the changing room means changing out is unpaid; one before it means it is not.
But the end adds a wrinkle: people frequently punch out on the way past and then return for something. The record then shows an earlier finish than the exit, and the organisation has a record it cannot explain.
Putting the terminal at the exit, so that the punch and the departure are the same act, solves both ends of this and is the single most common recommendation in this collection.
The waiting case
A worker who has finished, has punched out, and cannot leave — because the door is locked, the bus is in twenty minutes, or the site requires an escort — is in an awkward category.
If the constraint is the employer's arrangement and the person is not free to go, the position in several jurisdictions is that the time is still working time. If they are free to leave and choose not to, it is not.
The practical point is to remove the constraint rather than argue about it. A door that unlocks at the shift end costs nothing.
Measuring both ends together
The useful figure is the edge for the whole day: entry to punch, plus punch out to exit.
Reporting only the start understates by a third in most operations, and it gives anybody who wants to dismiss the finding an easy route — the figure is incomplete and they can say so.
What to change first
Move the terminal to the exit. Unlock the door at the shift end. Stop querying punch-outs past the hour, and watch what happens to the cluster.
Those three cost almost nothing and between them address most of what this note describes.
The last person out
On most sites one person does more at the end than everybody else: locks up, sets the alarm, checks the floor, waits for the last vehicle. It is frequently the supervisor and it is frequently unrecorded, because they are salaried and nobody thinks of it as a shift.
Fifteen minutes every evening is a substantial annual figure and it sits with a single person who is in no position to raise it. Either the role is scheduled to include it, which means the nominal finish is later, or the tasks are distributed. Leaving it where it is relies on the goodwill of the one person whose goodwill the site can least afford to lose.