Why Your OEE Number Might Be Lying to You

A dashboard can make bad data look very convincing.

OEE for example is one of the most useful and high-level metrics in manufacturing.

It's also surprisingly easy to make look more accurate than it really is.

A factory might report an OEE of 82.4%.

It looks precise.

But how precise is it?

Was every downtime event recorded?

Were planned stops separated from unplanned downtime?

Is the “ideal” cycle time still accurate?

Were short micro-stops captured, or simply ignored?

Were scrap and rework attributed to the right machine?

And if different departments calculate OEE differently, which number are you actually looking at?

The calculation itself is simple:

OEE = Availability × Performance × Quality

The difficult part is making sure those inputs reflect what actually happened on the shop floor.

A machine might stop for 90 seconds, restart, then stop again two minutes later. If those micro-stops aren't recorded, its availability may look better than reality.

If the ideal cycle time hasn't been updated for years, performance can be distorted.

If operators enter downtime manually at the end of a shift, the result may depend on what they remember.

Before building another dashboard, ask whether the factory can trust the data feeding it.

Because a sophisticated KPI built on poor data doesn't create better decisions.

It just gives bad information a more professional appearance.