An airplane typically has nine nines of reliability. 99.9999999% reliable. If it was ONLY 99.99% I believe the statistic is there would be 4-5 plane crashes PER DAY! After all, that’s a 1 in 10,000 failure rate. At that reliability, no one would take air travel. It would feel too risky.
Each nine adds another factor of ten to the acceptable failure rate:
- 99.999% reliability is a failure of 1 in 100,000 – One crash every other day
- 99.9999% reliability is a failure of 1 in 1,000,000 – One crash every 3 weeks
- 99.99999% reliability is a failure of 1 in 10,000,000 – One crash every 30 weeks
- 99.999999% reliability is a failure of 1 in 100,000,000 – One crash every 300 weeks
- 99.9999999% reliability is a failure of 1 in 1,000,000,000 – One crash every 3,000 weeks
100% reliability is a dream world where nothing ever goes wrong. The best thing we can do is check that we are using the right number of nines for the work we do. What are the consequences when something goes wrong? The higher the consequences, the more nines you need. Of course, these are just targets. 3,000 weeks is over 57 years. In the history of air travel, there has been more than 2 accidents, so that rate isn’t holding up, but these standards also didn’t exist for the entire history of air travel, they were developed over time. Earlier on, there were less sophisticated models of reliability and we had to learn.
Standards like this can be applied to more than safety. You could apply this to a restaurant and the taste of the dishes being made. You could apply this to a call center and the time it takes them to answer the phone. You could apply this to a website and the amount of uptime it has.
Have you thought about the number of nines you need in your industry in order to stand out?