Managing Overtime Fairly Without Losing Track

Overtime is the most honest number in a company. It tells you where the work really is, which manager cannot say no, and which team is running on goodwill that is about to run out.

Most SMEs only look at it once a month, when payroll needs a figure, by which point it is a cost to be paid instead of a signal to act on.

Record it the same way every time

The first problem is usually not the amount. It is that overtime gets recorded three different ways in the same company. One team logs it in the attendance system. One manager keeps a note on their phone. One department has an unwritten arrangement where extra hours become time off later, which appears in no record at all.

Pick one method and use it for everybody, including the people who are senior enough that nobody usually asks. Informal arrangements feel generous until the person leaves and wants their hours paid out, and nobody can prove what was agreed.

Decide the approval rules before the busy month

Rules written during a peak are not rules, they are permissions. Write them in a quiet week.

• Who can authorise overtime, and up to how many hours without escalating
• The monthly ceiling per person, and what happens when someone reaches it
• Whether it is paid, compensated with time off, or the person chooses
• How much notice staff get before being asked to stay late

That last one gets forgotten and matters more than the rate. Being asked at 4pm to stay until 8pm is a different thing from knowing on Monday that Thursday will be long.

Read the pattern, not the total

Once you have three months of consistent data, sort it by person and by team, and look for shapes.

The same three names every month is not a busy period. It is a staffing gap, or a manager who delegates to whoever says yes. One team consistently at double the company average usually means the workload was distributed badly a year ago and nobody revisited it. And a person whose overtime suddenly drops to zero after months at the top is often a person who has mentally left.

I would treat sustained high overtime as a resignation risk before treating it as a cost problem, though the cost problem is real too.

One number to watch

Overtime as a percentage of total hours, per team, per month. Not the peso amount, which moves with salary changes and hides the trend.

If that percentage is stable and low, your planning is working. If it climbs for two months in a row, something structural changed and the overtime is just the first place it showed up.

TalentHR ties the approval, the recorded hours and the payroll line together, so overtime stops being a monthly reconciliation and starts being something you can see while it is happening. That is the version worth having.