An 85% OEE can sit on top of six figures of recoverable loss. The metric isn't wrong — it's just averaged to the point of being useless for action. Read it like a dashboard and you'll feel fine. Read it like an operator and you'll find the money.
The Trap
OEE is a product of three things — availability, performance, and quality. Multiply them and you get a single, comforting percentage. The trouble is that the same 85% can come from wildly different plants: one bleeding availability to changeovers, another bleeding performance to micro-stops, a third bleeding quality to rework.
Each of those needs a completely different fix. The headline number tells you none of it. A metric that can't distinguish between those three plants isn't a diagnosis — it's a mood.
The Hidden Losses
When we decompose OEE against real plant data, the recoverable money almost always hides in losses that the top-line number smooths over:
The two-minute stoppages no one logs. Individually invisible, collectively the single biggest performance leak on most lines.
Equipment quietly running below its rated rate because 'that's just how it runs.' A slow line looks busy and reports fine.
Availability lost to changeovers treated as fixed cost rather than a target — often the fastest place to recover hours.
Rework and scrap that get absorbed into 'normal' yield and never surface as the OEE penalty they actually are.
The Discipline
The point of measurement is to change a decision. If a number can't tell an operator what to do differently this shift, it's overhead. Nextep's assessments don't stop at the OEE figure — we decompose it, tie each loss to a dollar value, and rank it by how hard it is to capture. That's the difference between 'we should improve' and 'here is exactly where, worth exactly this much, in this order.'
From Metric to Money
Once each loss carries a dollar figure, the conversation changes. Priorities stop being a debate about opinions and become a decision about returns. The micro-stops you couldn't see are suddenly the highest-ROI project on the floor.
OEE is still worth tracking. Just don't mistake the average for the answer. The answer is always one level down.