The input that decides whether performance means anything.

Not the quoted rate, not the average, not the theoretical maximum. What ideal cycle time is, how it goes wrong, and what to do when you do not have one.

Work your own numbers → WhatsApp us

Ideal cycle time is the input that decides whether an OEE performance figure means anything, and it is the input most plants cannot defend when questioned. It is also the reason the MachineWise variant exists: a metric that cannot be computed until a standard is agreed is a metric that never gets computed on a high-mix floor.

This article sets out what ideal cycle time actually is, the four ways it goes wrong, and what to do when you do not have one.

What it is, precisely

The ideal cycle time is the fastest the machine has demonstrably produced that part under normal conditions. Three things it is not, each of which is commonly substituted for it:

Not the quoted rate. The figure used to price the job was an estimate made before the job ran, often deliberately conservative, and using it makes the machine look better than it is.

Not the average. An average includes every slow cycle, so performance computed against it flatters you by construction and can never fall below roughly 100%.

Not the theoretical maximum. A rate the machine could achieve with a tool it does not have on material it does not run is not a standard, it is an aspiration, and performance measured against it is permanently poor for no useful reason.

The four failure modes

FAILURE 01

Set once at quotation, never revisited

The most common. Tooling, materials and programs change; the standard does not, and the performance figure gradually detaches from reality.

FAILURE 02

Optimistic, to win the job

A standard set to justify a price produces a permanently poor performance figure, which teaches everyone to ignore the metric.

FAILURE 03

One standard for a family of parts

Convenient and wrong. Variants within a family genuinely differ, and averaging them means every individual figure is incorrect.

FAILURE 04

Standard includes handling time

If the standard covers load and unload but the measurement covers only cycle time, or vice versa, the mismatch is invisible and consistent.

How to establish one you can defend

OBSERVEMeasure real cyclesCollect actual cycle times for that part over a representative period, not a single good run.

SELECTTake the best sustained rateThe fastest cycle achieved repeatedly under normal conditions, not the single quickest instance.

DOCUMENTRecord how it was derivedDate, machine, tooling, material and operator. A standard whose origin is unknown cannot be defended.

REVIEWRevisit on changeWhenever tooling, program or material changes. A standard is a living figure, not a constant.

This is unglamorous and it is the difference between a performance figure that survives scrutiny and one that is quietly ignored. Where a monitoring system records cycle-time distributions automatically, the first step is already done — the data is there, and the standard becomes an observation rather than a negotiation.

What to do when you have no standard

On a floor running fifty part numbers a month, establishing and maintaining ideal cycle times for all of them is a real project, and the honest answer is that many plants will not do it. Waiting until they have is how plants go years without any OEE at all.

The practical route is to use the MachineWise variant in the meantime. Substituting productive time ÷ machine on-time for the speed question needs no standard, is directly observable, and gives you a usable figure from day one. Its limitation is stated plainly: it cannot detect a machine running slowly, because a machine cutting continuously at half rate still counts as productive.

So the sensible sequence is to start with the variant across the floor, then establish proper ideal cycle times for your highest-volume parts — where the speed question is worth real money — and compute ISO figures for those. Both are available side by side in the OEE calculator.

Questions

Straight answers.

What is ideal cycle time?

The fastest the machine has demonstrably produced that part under normal conditions. It is not the quoted rate, not the average, and not a theoretical maximum.

Why not use the quoted cycle time?

Because it was an estimate made before the job ran, often deliberately conservative. Using it makes the machine look better than it is and makes the performance figure indefensible.

Why not use the average cycle time?

Because an average includes every slow cycle, so performance measured against it flatters you by construction and can barely fall below 100%.

How often should ideal cycle times be reviewed?

Whenever tooling, program or material changes. A standard set at quotation and never revisited is the most common cause of a meaningless performance figure.

What if we have no ideal cycle times at all?

Use the MachineWise variant, which substitutes productive time over machine on-time and needs no standard. Then establish proper standards for your highest-volume parts, where the speed question is worth real money.

What is the limitation of the time-in-cut substitute?

It cannot detect a machine running slowly. A machine cutting continuously at half its rated speed still counts as productive, which is exactly what the ISO effectiveness measure would catch.