Rounding, and Where the Minutes Go

Nobody works in neat units and every system records in them. The gap between the two is resolved by a rounding rule, and the rule is usually whatever the product shipped with.

The record described in “Rounding, and Where the Minutes Go” should be created close enough to the work that people are not reconstructing a polished week from memory. When teams assess a practical route to time tracking with screenshots for time tracking with screenshots, they should keep entry, project selection and correction simple, while explaining which optional activity data is collected and how employees can review it.

The amounts look trivial per entry and are not trivial in aggregate. Eight minutes a day across two hundred people is a little over twenty-six full-time days a year, moving in one direction. Which direction depends on a configuration setting that, in most organisations, nobody chose.

For a separate benchmark relevant to “Rounding, and Where the Minutes Go”, consult the WHO mental-health-at-work resources. 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.

The three decisions

The increment: one minute, six minutes, a quarter hour, half an hour. Six minutes is the traditional professional-services unit because it divides an hour into ten; fifteen is the most common in workforce systems because it matches shift patterns.

The direction: up, down, or to nearest. To nearest is the only one that is symmetric, and it is not the default everywhere.

The level: rounding each entry, each day, or the week. This is the decision people miss and it has the largest effect. Rounding six entries up to the quarter hour adds up to ninety minutes a day. Rounding the daily total adds up to fifteen. Rounding the week adds at most fifteen for the whole week, which is why rounding at the latest possible point is almost always the right answer.

Where it becomes a legal question

In several jurisdictions, rounding that systematically favours the employer is unlawful regardless of how small the individual amounts are, and the test applied is whether the practice is neutral over time rather than whether any single instance was fair.

The pattern that attracts attention is asymmetric rounding at the clock: rounding start times forward to the next quarter and end times back to the previous one, which takes up to twenty-eight minutes a day from every shift. It is a default in some time-clock configurations, it is indefensible, and it produces back-pay liabilities that dwarf whatever it saved. Check it before anything else.

Where it becomes a billing question

Where time is billed, the rounding rule is usually set by the engagement terms rather than by the system, and the two frequently disagree.

A firm whose standard terms specify six-minute increments and whose system rounds each entry up to the quarter hour is over-billing by rule, and the client's auditors will find it. The reverse — a system rounding to the minute against terms that permit six-minute units — is leaving money on the table, which is less dangerous but equally unintended. Reconciling the configuration against the terms is an hour of work and it is almost never done after the initial implementation.

Minimums and their effect

Separate from rounding but frequently confused with it: minimum charge units. A fifteen-minute minimum per task means a two-minute phone call is recorded as fifteen minutes.

This is defensible where it reflects a real cost — the interruption genuinely costs more than the call — and indefensible where it is applied to internal cost allocation, because it inflates the apparent cost of the activity and distorts every decision made on the basis of it. If you have minimums, know where they apply, and do not let them leak from billing into internal reporting.

The check worth running

Take a period's raw clock or entry data and the same period's rounded output, and compute the difference in total hours, by person and by direction.

If the net effect is close to zero and the per-person spread is symmetric, the rule is neutral and you can stop. If the net is consistently positive or negative, you have a systematic transfer, and you now know its size. If the per-person spread is wide — some people consistently gaining, others consistently losing — the cause is usually shift patterns interacting with the increment, and it is worth understanding before somebody notices it independently.

What to write down

The increment, the direction, the level at which it applies, and the date the configuration was last changed. One short page.

It is the document that answers the question when it is asked, and the question is always asked at the worst moment: in a pay dispute, in a client audit, or on the day somebody in payroll works out what twenty-six days a year is worth.