OEE vs Performance: Why Performance Is Usually the Hidden Loss

Key takeaways

• Performance is one of three OEE factors — it measures actual speed vs ideal cycle time.
• It is the factor most plants under-instrument because it requires a reliable ideal cycle time per SKU.
• Micro-stops (pauses under the downtime threshold) and slow cycles are the two biggest hidden Performance losses.
• A line at 95% Availability and 95% Quality with 70% Performance is doing 63% OEE — and the loss is invisible to anyone watching only uptime.
• The fix usually requires line-side cycle-count instrumentation, not just shop-floor reporting.

Short answer: Performance is the OEE factor that measures whether the machine is running at design speed. Most plants get Performance wrong because they do not have a reliable ideal cycle time per SKU, or they cannot detect micro-stops smaller than the downtime threshold. Performance loss is usually the biggest under-measured OEE component and the place where the most uncaught capacity sits. See also OEE vs Utilization.

What Performance measures

Performance is the ratio of theoretical maximum output to actual output during run time.

Performance = (Ideal cycle time x Total parts produced) / Run time

If ideal cycle time is 30s, you produced 600 parts in 432 minutes of run time, then Performance = (30 x 600) / (432 x 60) = 18000 / 25920 = 69.4%.

The other equivalent form: Performance = Actual run rate / Ideal run rate.

Why Performance is the hardest factor to measure

Three things have to be right:

• Ideal cycle time per SKU. The design speed for the part being made, not a generic line speed. Many plants do not maintain this number per SKU and use a line average — which makes Performance noise.
• Accurate part counts. From PLC, vision, or operator entry. Operator entry is the least accurate.
• Honest run time. Anything not counted as downtime becomes run time in the denominator. Plants that under-log micro-stops inflate run time and crush Performance.

Plants that report Availability and Quality but skip Performance usually do so because one of these three is broken.

The two big hidden Performance losses

Micro-stops. Pauses of 1-60 seconds that are too short for the downtime system to log. They show up as a Performance loss because run time is intact but throughput drops.

Slow cycles. The machine is running but not at design speed. Worn tooling, suboptimal recipe parameters, or just operator habit can cost 10-20% of design speed without anyone noticing.

A 95% Availability line with frequent micro-stops and slow cycles might run at 70% Performance. That is 25 points of OEE loss invisible to a downtime-only system.

What hides Performance loss

• Operator-reported part counts. Operators round and batch their entries, which smears the cycle-time signal.
• Generic ideal cycle times. Using one number for the whole line instead of per-SKU makes Performance unreliable.
• Downtime thresholds that are too high. If anything under 60 seconds is ignored, all the micro-stops disappear into run time.

How to surface Performance loss

1. Capture cycle counts from PLC or vision. Operator entry is too coarse.
2. Maintain ideal cycle time per SKU. One number per part, kept current as tooling and recipes change.
3. Drop the micro-stop threshold. Log pauses down to 5-10 seconds so micro-stops show as downtime, not as a Performance ghost.
4. Reconcile every shift. If reported parts and PLC cycle count disagree by more than a few percent, the data is wrong somewhere.

Why Performance is high-leverage

Availability losses are usually known — the line was down, everyone saw it. Quality losses are mostly known too — scrap goes in a bin. Performance is the only OEE factor where the loss can be invisible to everyone on the shop floor. Lighting up Performance is usually the single highest-leverage move an OEE platform can deliver.

Fabrico's OEE module captures cycle counts from PLC and lets you set ideal cycle time per SKU, so Performance reads true. The output is a list of the lines where Performance is the dominant loss, not an aggregate average that hides which line needs work.

See how Fabrico captures this automatically — explore OEE for manufacturing or book a demo.

Related reading

• OEE vs Utilization
• OEE and ISO 22400
• OEE for Batch Manufacturing
• OEE vs Availability

Frequently asked questions

Is Performance the same as speed?

Roughly yes. Performance is actual rate as a percentage of design rate. "Speed loss" is the common shop-floor term for the same thing.

What is a good Performance target?

World-class is around 95%. The 85% OEE world-class benchmark assumes Performance ~95%, Availability ~90%, Quality ~99%.

Can Performance be greater than 100%?

Yes, if the ideal cycle time is set too slow. That is a signal to update ideal cycle time, not to celebrate.

Why is my Performance always low?

Either ideal cycle time is wrong (often too aggressive), or micro-stops and slow cycles are eating real capacity. Look at the cycle-time distribution per SKU to tell which.

How is Performance different from utilization?

Utilization is run time vs scheduled time (close to Availability). Performance is actual speed vs design speed during run time.