Overall equipment effectiveness is widely reported and frequently misused. It gets presented as a single
number, tracked as a target, and improved in ways that don’t increase output. The metric is sound; the way
it’s usually applied isn’t.

The Three Components

OEE multiplies availability, performance, and quality.
Availability is the share of planned production time the equipment was actually running — losses here are
breakdowns, changeovers, and waiting for material or operators.
Performance is actual output rate against the designed rate. Losses are slow running and short stops that
don’t get recorded as downtime.
Quality is the share of output that was right first time. Losses are scrap and rework.
Multiplying them means the composite falls fast. Three components each at ninety percent give roughly
seventy-three percent overall, which is why OEE figures look low compared to any individual measure.

The Mistake That Makes It Useless

Reporting a single plant-wide OEE. It averages together equipment that limits your output and equipment
that doesn’t, producing a number that can improve while output stays flat.
OEE on a non-constraint machine is close to meaningless. If that machine is idle half the time because it’s
waiting for the bottleneck, its OEE is poor and nothing is wrong. Improving it produces more work in
progress and no more finished goods.
Measure OEE on the constraint. There, every lost minute is a lost minute of plant output, and improvement
translates directly. Elsewhere, track availability for maintenance purposes and don’t manage to the
composite.

Definitions That Get Gamed

Planned production time is the classic. Exclude enough as planned downtime — changeovers, breaks,
planned maintenance, slow periods — and OEE rises without anything improving. Be consistent, document
what’s excluded, and be suspicious of a step change in the number that coincides with a definitional
revision.
Ideal cycle time matters too. Set it to what the machine actually achieves and performance sits near a
hundred percent permanently, telling you nothing. Set it to the manufacturer’s rate and you have a real
reference.

Using it Well

Report the three components separately alongside the composite. The composite tells you how much is
lost; only the components tell you where, and they lead to completely different actions. Availability losses

Go to maintenance and changeover work. Performance losses go to process and tooling. Quality losses go
to root cause analysis.
Track loss reasons underneath availability with a short, consistent code list. The Pareto distribution of
reasons is more actionable than the OEE figure itself.
And put the number in context. OEE targets borrowed from other industries or from published world-class
figures are close to meaningless across different processes and product mixes. Your own trend is the
comparison that means something.
Be careful about tying incentives to OEE as a headline figure. Where bonuses depend on it, definitions get
reinterpreted, short stops go unrecorded, and the number improves while output doesn’t. Incentivise
output and cost; use OEE as the diagnostic that explains them.
ticktick.ai calculates OEE by work centre with the components and loss reasons broken out, and identifies
which equipment is currently constraining output.