The Schedule-to-Time Join

The rota says what should happen. Time capture says what did. Comparing them is where variance analysis lives and where most of the noise comes from.

What crosses

Outbound: the schedule. Who, where, when, which role.

Inbound: the actual. Punches, or a marked-up rota, or nothing at all where attendance is assumed.

And the difference is the finding: someone worked a shift they were not scheduled for, or did not work one they were.

The four disagreements

Unscheduled work. Someone covered and nobody updated the rota. Common, and the hours are real.

Scheduled and absent, which needs an absence reason from a third system.

Different times. Scheduled nine to five, worked eight-thirty to five-fifteen — which is a rounding, overtime and tolerance question rolled into one.

Different person. A swap that happened and was never recorded, so the hours attach to the wrong name.

Which system is right

Neither, and that is the point.

The schedule is a plan; the clock is an observation.

Where they disagree, the clock is usually closer to reality — but not always, because clocks fail and swaps are real.

The rule has to be stated: what gets paid when they differ, who decides, and by when.

Most organisations have never written it down, and the answer is therefore whatever the payroll clerk did last time.

Tolerance

A small difference should not require a decision.

State the tolerance — a few minutes either side — and let anything inside it pass without review.

Anything outside it goes to an exception list that a human looks at.

Without a tolerance, every shift is an exception, and the review becomes rubber-stamping.

What to measure

Exception rate: shifts where schedule and actual differ beyond tolerance, as a proportion.

By site, by week.

A site far above the others has a rota problem, a clock problem or a swap-recording problem, and the three are distinguishable by looking at which of the four disagreements dominates.

And the trend, because a rising exception rate means the rota is describing reality less well each week.

Set the tolerance deliberately

Otherwise every shift is an exception.

A few minutes either side, passing without review.

Anything outside it to a list a human reads.

Without a tolerance the exception list is the whole rota, and the review becomes rubber-stamping — which is worse than no review, because it looks like a control.

A practical configuration prompt

During configuration, use remote employee monitoring platform to prompt questions about identifiers, ownership and output. Treat the page as a starting point and document each assumption.