Process Capability
5 min read · 4 principles · apply it by lunch
Good morning. Your average is fine. Your customer never experiences the average.
You promise your customers something with a number in it — delivery in five days, a reply within the hour, an error rate under one per cent — so what proportion of the time do you actually hit it, measured rather than felt? If the only figure you’ve ever reported is the mean, the answer is that you don’t know, and neither does the customer who has been quietly counting the misses.
In 1986 Bill Smith, an engineer at Motorola, made an argument that grew into Six Sigma: products needing rework on the line failed disproportionately early in the customer’s hands, so the defects you catch internally forecast the ones the customer will find. Acting on that requires a way to set how much your process varies against how much variation the customer will tolerate, and the arithmetic is disarmingly simple. Put two widths side by side: the span your process ordinarily covers, which is six standard deviations of its own results, and the tolerance your customer will actually accept. Divide the second by the first and you have Cp. If Cp is 1.0 the two are the same width, your specification sits exactly three standard deviations either side of centre, and roughly 2,700 units in every million fall outside it — and that assumes perfect centring, which no process has. Plenty of industrial customers won’t accept anything under 1.33. You don’t need the full statistics to use any of this: count your last thirty results, count how many fell outside the line the customer will actually take, and you’ve got the same number in cruder form — which is more than most businesses can produce for any promise they’ve already signed.
Process Capability is Day 345 of 365.
In the Daily, each of its four principles is explained in full, with a one-page template to apply it the same day. A new framework arrives every morning at 7am for a year, across all fifteen courses of the FutureMBA curriculum.
$29 a month · cancel any time · 7-day money-back guarantee