OEE in practice: what the numbers actually mean
OEE — Overall Equipment Effectiveness — is the product of three factors:
OEE = Availability × Performance × Quality
Each factor is a ratio between 0 and 1 (or 0% and 100%). A “world-class” OEE for discrete manufacturing is typically cited as 85%. In practice, most operations run between 40% and 70%, and many have no idea what their actual number is.
This post isn’t about how to calculate OEE. It’s about how to read OEE numbers honestly — what they tell you, what they hide, and how to use them to make decisions.
What each factor actually measures
Availability measures how much of your planned production time was actually spent producing — versus being lost to downtime (breakdowns, changeovers, material shortages, maintenance, etc.).
A machine with 90% availability loses one hour in every ten to planned or unplanned stoppages. If you run two shifts of eight hours each, that’s 1.6 hours of lost capacity per day.
Performance measures how fast you ran versus your standard rate. A machine running at 80% performance is producing at 80% of its rated speed — either because the machine is cycling slowly, the operator is cycling slowly, or there are micro-stoppages (brief pauses under a minute each) that add up.
Performance below 85% usually means one of three things: the standard rate is wrong (too aggressive), there’s a recurring micro-stoppage that nobody has investigated, or the operator workflow is creating pace drag.
Quality measures the proportion of output that meets spec on first pass, without rework or scrap. A 95% quality rate means 5% of your parts are either scrapped or sent to rework — which means you’re doing more actual cycles than your schedule requires.
The compounding effect
Here’s what makes OEE deceptive: the factors compound.
A machine running at:
- 90% Availability
- 90% Performance
- 98% Quality
…has an OEE of 79.4%. That seems reasonable. But if you push Availability to 95%, Performance to 95%, and Quality to 99.5%, OEE jumps to 89.5%. The delta isn’t additive — each improvement multiplies.
This means chasing a single factor to 100% while ignoring the others is inefficient. The highest-leverage improvements are usually the ones that raise the lowest factor.
What OEE doesn’t tell you
OEE doesn’t tell you whether your equipment is the bottleneck. A 60% OEE on a feeder machine doesn’t matter if the bottleneck is downstream. Goldratt’s Theory of Constraints applies: improve the bottleneck first.
OEE doesn’t tell you why. It’s a symptom metric, not a diagnostic. When Availability drops from 88% to 72%, something changed — but OEE alone doesn’t tell you it was a bearing failure on machine 3 or a shift changeover that ran 45 minutes instead of 15. You need the underlying downtime events and their reasons to diagnose root cause.
OEE doesn’t tell you whether your standard rates are realistic. Performance above 100% is theoretically possible (if a machine runs faster than its standard rate in some conditions) and is a red flag that the standard is wrong, not that you’ve achieved superhuman efficiency.
The “world-class 85%” benchmark is misleading
The 85% figure comes from early Lean Manufacturing literature focused on high-volume, low-mix automotive production. For a job shop running 200 different part numbers on 15 machines, 85% may be aspirational in one direction or the wrong benchmark entirely in another.
Better: compare your OEE against your own historical baseline, segment by equipment type, and track trends over time. A job shop improving from 52% to 61% OEE over six months is making real progress — regardless of what a benchmark says.
How to use OEE operationally
The most valuable OEE practice is shift-level visibility, not monthly reporting. A supervisor who sees real-time OEE on a floor display can respond to a Availability drop in the same shift it happens. A plant manager who sees monthly OEE in a PDF report is reading history.
Effective OEE-driven operations typically:
- Capture downtime reasons at the machine level (ideally at the minute of occurrence, not shift-end)
- Review shift OEE in a daily standup with operators present
- Use Pareto analysis on downtime reasons to identify the top 3 losses per equipment per week
- Run focused improvement events (kaizens, SMED workshops) on the top losses
The data is only useful if it changes behavior. OEE as a reporting exercise doesn’t improve yield.
Qontiv tracks OEE automatically from machine signals and downtime events — no manual data entry. See how it works or start free.