Growth Marketing Glossary

Process Capability (Cp and Cpk)

cee·pee cee·pee·kaynoun

Can the process hit the target, reliably? Cp asks whether the process is tight enough to fit the limits; Cpk asks whether it is both tight enough and centered.

process variationmeasure capability vs limitsCp and Cpk
Schematic — process spread and centering against limits
Term
Process capability (Cp and Cpk)
Are
Six Sigma capability indices
Cp measures
Spread versus tolerance width
Cpk adds
How well the process is centered

Parts of speech & senses

process capability · noun
  1. Cp and Cpk are process capability indices used in Six Sigma. Cp measures how well a process's spread fits within its specification limits, and Cpk also accounts for how well the process is centered. "Cp looked fine, but poor centering dragged Cpk down."

What Cp and Cpk are

Cp and Cpk are process capability indices — numbers used in Six Sigma and quality engineering to judge whether a process can reliably produce output within its specification limits. Cp, the capability index, compares the width of the allowed tolerance (the gap between the upper and lower specification limits) with the natural spread of the process, defined as six standard deviations. Its formula is the tolerance width divided by six times the process standard deviation: Cp = (USL − LSL) / (6σ). A Cp of 1 means the process spread just fills the tolerance; a Cp above 1 means the spread is comfortably narrower than the limits, which is what you want. But Cp has a blind spot: it assumes the process is centered between the limits and only measures whether it is tight enough, ignoring where the process is actually aimed.

Cpk closes that blind spot by adding centering. It measures the distance from the process mean to the nearest specification limit, in units of three standard deviations, and reports the worse of the two sides. So Cpk asks not just whether the spread fits, but whether the process is both tight enough and aimed at the right place. If a process is perfectly centered between its limits, Cpk equals Cp; if it drifts off center, Cpk falls below Cp, because one side is now closer to going out of specification. Cpk is therefore always less than or equal to Cp. Together the two tell a fuller story than either alone: Cp is the potential capability if the process were centered, and Cpk is the actual capability given where it currently sits.

Cp versus Cpk

The cleanest way to hold the difference is an analogy quality engineers use: Cp is the size of your car relative to the garage, while Cpk is your ability to actually park it inside. Cp tells you whether the car is narrow enough to fit through the doorway at all — whether the process spread is tight enough for the tolerance. Cpk tells you whether you have also aimed it correctly, or whether you are scraping one side of the garage because the process is off center. A process can have an excellent Cp — plenty narrow — yet a poor Cpk, because it is running consistently high or low and one side is nearly out of specification. That gap between a good Cp and a worse Cpk is the signal that the process is capable but poorly centered, and needs re-aiming rather than tightening.

This distinction drives different fixes. When Cp itself is low, the process is too variable — its spread is too wide for the tolerance — and the remedy is to reduce variation: tighter control, better equipment, less noise in the inputs. When Cp is healthy but Cpk lags, the process is precise but off target, and the remedy is to shift the mean back to center, which is often a simpler adjustment than reducing variation. Reading both together tells you which problem you have. Reporting only Cp flatters a process that is quietly drifting toward a limit; reporting only Cpk can hide how much headroom the process would have if it were re-centered. The pair, read side by side, separates the two independent questions of spread and aim.

Using Cp and Cpk well

Using Cp and Cpk well starts with computing them honestly from a stable, in-control process, because both indices assume the data are roughly normal and the process is stable over time — feed them an unstable process and the numbers are meaningless. Report the pair together, not one alone, so a reader can see both spread and centering. Use the gap between them diagnostically: a healthy Cp with a low Cpk points to a centering problem to be fixed by re-aiming the mean, while a low Cp points to excess variation to be reduced at the source. Set capability targets appropriate to the stakes — higher indices for critical characteristics — and remember that a Cpk of 2 corresponds to the classic six-sigma level of capability, a demanding standard reserved for the characteristics that truly warrant it.

The failures start with computing the indices on an unstable process, where the standard deviation is not meaningful and the numbers mislead. People report Cp alone and miss that a well-behaved-looking process is drifting toward a limit, or report Cpk without Cp and lose sight of how much potential a re-centering would unlock. They assume normality that the data do not support, chase an ever-higher index on a characteristic that does not need it, or treat a single snapshot as proof of ongoing capability when the process can drift. The discipline is to stabilize first, check the distribution, report Cp and Cpk together, read the gap between them to decide whether to reduce variation or re-center, and match the target to how much the characteristic actually matters.

Worked example. A parts line has a tight, consistent process, and its Cp comes out well above 1 — the spread easily fits the tolerance, so on paper the process looks highly capable. But the finished parts keep failing inspection on one side. Computing Cpk reveals the problem: the process is running consistently high, off center toward the upper limit, so Cpk is far below Cp even though the spread is fine. The fix is not to reduce variation but to re-aim the mean back to center, after which Cpk rises to meet Cp and the failures stop. Cp alone would have hidden the drift; the pair, read together, pointed straight at the cause. (Illustrative; RGM analysis.)
Failure modes to watch. Computing the indices on an unstable, out-of-control process where the standard deviation is meaningless; reporting Cp alone and missing a process drifting toward a limit; reporting Cpk without Cp and losing sight of re-centering potential; assuming normality the data do not support; and chasing an ever-higher index on a characteristic that does not need it.

Synonyms & antonyms

Synonyms

process capability indexCpkcapability ratio

Antonyms

process incapabilityuncentered process

Origin & history

Cp and Cpk — Six Sigma process-capability indices — measure whether a process fits its limits (Cp) and whether it is also centered (Cpk), so Cpk is always at most Cp.

Etymology: source.

Usage trends

Search interest for this term over the last five years:

View interest-over-time on Google Trends →

Common questions

What are Cp and Cpk?
Process capability indices used in Six Sigma. Cp measures how well a process's spread fits within its specification limits, calculated as the tolerance width divided by six standard deviations. Cpk adds how well the process is centered on target.
What is the difference between Cp and Cpk?
Cp measures only spread, assuming the process is centered — it asks whether the process is tight enough. Cpk also measures centering, so it reflects actual capability. If the process is perfectly centered, Cpk equals Cp; otherwise Cpk is lower.
Why is Cpk always less than or equal to Cp?
Because Cpk penalizes off-center processes. Cp assumes perfect centering and measures only spread. Cpk measures the distance from the mean to the nearest limit, so any drift off center lowers it below Cp. Only a perfectly centered process has Cpk equal to Cp.

Resources & people to follow

Curated, non-competitor resources verified per term.

Related training

Disciplines

Areas of marketing where process capability (cp and cpk) is a core concern:

Sources

  1. trendsGoogle Trends — "process capability index"