How do you measure warehouse labour utilisation?
Warehouse labour utilisation is the share of paid hours your people spend on the work, measured in time rather than floor space. Divide the hours spent on tasks, recorded and off-system, by the hours paid over the same shift, then multiply by 100. Measure it by zone and job, because a site-wide figure hides where the time goes.
The formula
Labour utilisation compares two things over the same shift or week: the hours your people spent on the work and the hours you paid them for.
Labour utilisation = hours spent on the work ÷ hours paid × 100
Hours spent on the work means every task the shift needed, whether the WMS recorded it or not: the picks, putaways, receipts and loads it does record, plus the replenishment, unloading, housekeeping and training it often doesn't. Hours paid come from the clocking or time and attendance system, from the moment somebody clocks on to the moment they clock off.
Both halves are simple to state. The trouble comes from what goes into each one.
Where it goes wrong: counting only what the WMS records
The WMS is the easiest place to find the top half of the sum, since it timestamps every task it hands out. It only sees the work it was set up to record, though, so the indirect jobs drop out and the figure reads low. In two warehouses we've worked in, the WMS put utilisation at around 42 per cent. With the off-system work counted, both read 65 per cent.
A figure that low tends to get read as people standing still, when most of the missing time went on work nobody logged.
Where it goes wrong: dividing by the wrong hours
The bottom half goes wrong the other way. When the hours come from the WMS as well, from first scan to last scan, the time between clocking on and the first task disappears, along with the time after the last one. The figure then reads high, because paid hours have dropped out of the sum without anybody noticing.
572 hours in one week sat between operators clocking on and starting their first task, time no report ever showed
The same team, measured three ways
Take a team of 20 on a 7.5-hour shift, which is 150 paid hours. The hours below are an example, set to land where two warehouses we've worked in did.
ONE TEAM, ONE SHIFT
The same work, divided three ways
Three bars for the same team and shift. WMS tasks only, over clocked hours, reads 42 per cent, which is low. All the work over clocked hours reads 65 per cent, which is right. All the work over first scan to last scan reads 75 per cent, which is high because paid hours drop out of the sum. A shaded band marks 85 to 90 per cent as realistic.
An example rather than a measurement, set to land where two warehouses we've worked in did.
| How it's worked out | Hours on the work | Hours divided by | Utilisation |
|---|---|---|---|
| WMS tasks only, over clocked hours | 63 | 150 | 42% |
| All the work, over clocked hours | 98 | 150 | 65% |
| All the work, over first scan to last scan | 98 | 130 | 75% |
An example rather than a measurement. The first two rows match what two warehouses we've worked in measured on their systems and then with the off-system work counted.
The third row reads ten points higher than the second while 20 paid hours a shift never reach the work. Only the second row is right. It needs the off-system work logged and the hours taken from the clocking system.
Utilisation isn't productivity
The two answer different questions. Utilisation asks whether the paid hours got to the work at all. Productivity asks how much got done once they did, as tasks an hour against what a reasonable operator does in the same job and zone. A warehouse can do well on one and badly on the other, which is why they're measured separately and read side by side.
Measuring it properly
- Take the paid hours from the clocking system rather than from the WMS.
- Have supervisors log the off-system work by activity: replenishment, unloading, housekeeping, training and loading.
- Work it out by zone, job and shift, since a site-wide figure averages the problem away.
- Stop once most of the day is accounted for. The last few minutes cost more supervisor time to capture than they tell you.
If nothing is logged yet, a day-in-the-life study shows which activities to log: somebody walks a shift with the people doing the work and writes down where the time goes.
Fettle works utilisation out this way, with the paid hours from the clocking system, the tasks from the WMS and the off-system work logged by supervisors, down to the zone and the shift.