Productivity is output divided by an input, such as units per labor hour. OEE is good output divided by what the equipment could ideally have made in its scheduled time. OEE can rise while labor productivity falls, as when you add a person to clear jams, so use the two together.
Two different questions
OEE asks how close the equipment came to its best during the time you planned to run it. It has a ceiling of 100%, and labor and cost never enter the sum.
Productivity asks how much output you got for what you put in. It has no ceiling, and it needs an input to divide by.
The US Bureau of Labor Statistics measures labor productivity as real output per labor hour. It also tracks a wider measure, output against combined inputs of hours, capital and purchases.
A plant usually works with simpler forms of the same idea.
What each one divides by
| OEE | Labor productivity | Machine productivity | |
|---|---|---|---|
| Formula | Good units ÷ (ideal rate x scheduled hours) | Units ÷ labor hours | Units ÷ machine hours |
| Unit | % | Units per hour | Units per hour |
| Upper limit | 100% | None | None, but capped in practice by the rate |
| Sees crew size? | No | Yes | No |
| Fair across products? | Yes, if each has its own ideal rate | Only within similar products | Only within similar products |
Three weeks where they split apart
Example: a packing line is scheduled for 40 hours a week, with an ideal rate of 600 units per hour. So it could ideally make 24,000 units.
Week 1 is the base case, with a crew of 3. Each of the next two weeks changes one thing.
| Week | What changed | Crew | Good units | OEE | Labor hours | Units per labor hour |
|---|---|---|---|---|---|---|
| 1 | Base case | 3 | 16,800 | 70.0% | 120 | 140 |
| 2 | One more person clears jams | 4 | 19,200 | 80.0% | 160 | 120 |
| 3 | Automatic feed replaces one job | 2 | 16,800 | 70.0% | 80 | 210 |
In week 2, OEE says things got better, and labor productivity says they got worse. Both are right, and the choice comes down to what 2,400 more units are worth against 40 more hours of pay.
In week 3, OEE sees no change at all, though the plant saved a full job. If you only tracked OEE, you would miss a real gain.
The bridge between the two
The two numbers are tied together by a short chain. It is worth knowing, because it shows exactly which lever you are pulling.
So you have three levers for output per person. Raise OEE, raise the ideal rate, or run with a smaller crew.
OEE measures only the first lever. A faster new machine changes the ideal rate, and then OEE may even fall on the first day while productivity climbs, as OEE after a new machine covers.
Why product mix fools productivity
Example: the same line runs a second product with an ideal rate of 300 units per hour, half the first. At the same 70% OEE, it makes 210 good units per scheduled hour instead of 420.
Units per hour have halved and nothing went wrong. OEE stays at 70%, so it shows that the line ran the same and the mix moved.
To fix this on the productivity side, use standard hours earned or value added instead of raw units. The guide to OEE with mixed products shows the OEE side.
When each one misleads you
- OEE misleads when labor is the main cost. A line at 80% OEE with too many people can lose money, and the score will not show that.
- OEE misleads when there are no orders. A low-demand month can leave OEE fine and productivity poor, as OEE with no orders explains.
- Productivity misleads when the mix changes, or when scrap counts as output. Count good units, or you reward making bad ones fast.
- Productivity misleads when stock builds up. Running flat out for the warehouse lifts output per hour and ties up cash.
How to use them together
Good units, scheduled hours and crew hours from the same shift record.
Look at the factors, then the biggest loss, to decide where to work.
Check units per labor hour and cost per unit before you make it permanent.
In week 2 above, the test is simple. Compare the pay for 40 extra hours with the margin on 2,400 extra units, and use the cost per unit guide for the sum.
Then ask whether a fix could get the same gain without the extra person. If those jams can be removed, you keep the 80% and return to 3 people, which gives 160 units per labor hour.
For a wider set of numbers, see production KPIs. If your plant reports efficiency against a standard, production efficiency vs OEE is the page you need.
Labor numbers usually exist already, while reliable stop and count data often does not. Fabrico takes stops, speed and counts from the machine and shows OEE per line and shift, so the two can be read side by side. See how Fabrico tracks stops.
What people ask next
Does a higher OEE mean higher productivity?
Usually, if crew and product stay the same, because more good units come from the same hours. It does not hold if you added people or overtime to get there. That is what week 2 above shows.
How do I calculate productivity from OEE?
Good units per scheduled hour = OEE x ideal rate. Divide by crew size for good units per labor hour, as long as the crew stays for all scheduled hours. Check the labor side with units per labor hour.
Is OEE a productivity measure?
It is often called one, but it has no input cost in it. It measures how well equipment used its scheduled time. Think of it as measuring one driver of productivity, not productivity itself.
Which should a plant manager report, OEE or productivity?
Both, for different audiences. OEE tells the floor where the losses are, and productivity or cost per unit tells finance what they are worth. See plant manager OEE KPIs.
What about overall labor effectiveness?
OLE applies a similar structure to people instead of machines, so it sits closer to OEE than plain units per hour does. For the plain labor number, see units per labor hour.
Fabrico is a light MES for OEE. It takes stops, speed and counts from PLCs or IoT sensors, with add-on AI cameras counting output where a machine has no usable signal, and operators tag the cause at the line.