How good is good?
Is 95% Good?
Sounds like an obvious question and if it was your exam result, you would probably be very happy with it. I know that I would have been very happy with that result, given some of the grades I actually got!! However, is it good in every environment?
I want to bring in some data from some other business segments I have worked in, namely Aerospace and Motorsport, and hopefully explain to you why 95% might not always be seen as an acceptable result.
At Heathrow Airport, official statistics show that on average they handle 1299 flights per day. If you take the assumption that success equals either a successful take-off or a successful landing, what does 95% success mean?
Based on the average of 1299 flights per day, this means that 65 planes PER DAY will not have a successful take off or landing i.e. CRASH
Is 95% good and acceptable? Would you ever get on a plane again?
One particular company is a major supplier of gearboxes into the motorsport market and in some of the formulas they are the sole supplier of gearboxes to every car/team in that particular formula.
Using Formula 1 as an example, in each race 20 cars line up on the grid and all of the race dates and times are known well ahead of each season. If the gearbox supplier was 95% successful it would mean that one car at every race would not have a gearbox in it when they lined up on the grid.
Is 95% good and acceptable? Would you want to be the driver without a gearbox?
You may think that is all well and good and very interesting but this is completely irrelevant to an education environment. If that is your thought pattern, then you may wish to re-consider when you read the following example.
I have recently been working on a tender for a Payroll and HR system where the Trust had approximately 2200 employees to pay. What struck me was the volume of payslips that would be generated over the course of the proposed contract i.e. 26,400 per year and 79,200 over the 3 year contract term. One of the potential suppliers was quoting a service level target of 95% for payroll accuracy which would mean that, if they achieved 95% accuracy, every month 110 employee payslips would be incorrect or to make the number slightly scarier 3960 payslips would be incorrect across the 3 year life of the contract.
Is 95% good and acceptable? Would you be happy if it was your payslip?
In the industries I have mentioned above (and many others) they have moved away from using percentages as a method of defining acceptable quality and now use measures such as 6 sigma and Defects Per Million Opportunities (DPMO). This makes it much easier to understand once you are looking at numbers which are over 90%. By converting from percentages to DPMO, 1% = 10,000 DPMO. Therefore 95% = 50,000 DPMO and using the payslips example this means that for every 1,000,000 payslips produced 50,000 of them will be incorrect. That is an awful lot of queries to resolve……….
Industries such as Aerospace use 6 sigma which is a way of easily assessing one company against another and each of the 6 levels equate to a different rate of DPMO & Percentage as below:-
1 sigma = 691,462 DPMO = 30.85%
2 sigma = 308,538 DPMO = 69.146%
3 sigma = 66,807 DPMO = 93.319%
4 sigma = 6210 DPMO = 99.379%
5 sigma = 233 DPMO = 99.9767%
6 sigma = 3.4 DPMO = 99.9997%
As you can see, even 4 sigma is well above 99% success. I don’t believe any manufacturing company in the world is genuinely consistently operating at 6 sigma, but some are getting close.
How good is good is a very subjective and situation dependent concept. With a bar of Cadbury’s wholenut chocolate I would be happy if only 80% of the squares actually had a nut in them, as long as the bar was generally full of nuts. However, I really want my payslip to be in the 6 sigma region and if I am ever to get on a plane again………………. I really do want them to have successful take-offs and landings way more than 95% of the time.
Steve Parker is Procurement Account Manager (Consultant) at Minerva Procurement Consultancy Services Limited.