An hour bank is a simple idea: extra hours worked do not go into the payslip, they go into an individual balance and come back as rest. Businesses like it because it smooths peaks with no immediate cost, and staff like it because it turns an extra Saturday into a Monday off.

And then, about half the time, it becomes a problem. Not because of how it is designed, but because of how it is kept.

What it is, technically

In the UK there is no statutory hour bank. What exists is time off in lieu, and in some businesses annualised hours contracts, both purely contractual arrangements. What goes in, at what rate, by when it must be taken and what happens to unused credit are all things your contract has to say, because nothing else says them for you.

That matters, because it means "how does an hour bank work" never has a generic answer. The rules below are about running it well, they do not replace what your contract states.

Elsewhere the equivalents are more codified: Italy has a contractual banca ore set out in sector agreements, and Spain combines irregular distribution of the working year with compensating overtime by rest within four months.

Why balances swell

When an hour bank goes wrong, it goes wrong the same way every time: the balance grows, nobody looks at it, and at some point it is too large to clear without leaving the floor uncovered.

The three causes, in order of frequency:

The balance is not visible to the person earning it. If the hours live in a file only the scheduler sees, nobody can ask to take them at the right moment, and they remember when it has become forty.

There is no real deadline. Credit without a date to be used by never gets used: it gets deferred to the quieter season, which arrives once the balance is already unmanageable.

There is no cap. With no per-person maximum, an hour bank stops being a flexibility tool and becomes a way of running overtime at deferred cost. That is the point where the relationship sours, because people understand the difference perfectly well.

The four rules that keep it healthy

Balance visible to the person earning it, always. Not on request, not at month end: on the phone, next to their shifts. It is the single measure that fixes more problems than all the others combined.

A declared window to use it. Three months, four, six, whatever your contract says. But declared, and with a reminder as it approaches.

A cap per person. Once it is passed, the conversation changes: either the time off gets scheduled or the hours get paid. It does not simply keep accruing.

Schedule the payback when the credit is earned, not when the credit is large. "You can take it back later" without a date is the most reliable way of never taking it back.

Keeping the count

The arithmetic itself is trivial: hours worked minus contracted hours, day by day, with a sign. The three places it goes wrong are elsewhere:

  • mixing overtime and banked hours in the same cell. They are two different destinies for the same extra hour, and the choice belongs at the moment, not at month end;
  • using planned hours instead of actual ones. A balance is built on the record. The method, including the minutes-to-decimals conversion that catches nearly everybody, is in how to calculate worked hours;
  • not keeping the history. A balance with no detail of how it formed is indefensible the first time it is questioned.

When not to use it

There is one case where an hour bank is the wrong answer: when the credit forms every week, on the same days, for the same people. That is not smoothing a peak, that is covering a headcount gap with a loan somebody will eventually want back.

How to tell the two apart is covered in overtime.

In Sked Solve the hours balance is visible to every team member in the app, next to their shifts, and the generator uses it to rebalance the load across following weeks. See how it works at Sked Solve.