The Algorithm (Do less to do more, vol. 4)

Walter Isaacson’s new biography of Elon Musk is OK-ish, but it’s full of practical advice (and also some madness; apparently, it’s a package deal). One of the best pieces (besides “hasve some money for one more rocket launch”) that fits right into my favourite "doing less with more" theme is called The Algorithm.

Elon and his teams use this 5-step process to get more done.

1. First, make the requirements less dumb. Do this by questioning why that thing is needed. This is by far the most important part. In every organization I’ve worked with, few dared to question requirements. Every requirement should have a single responsible person. Not a group, not a department - it’s difficult (and slow) to argue with those. For example, in banks and telecom companies, the words “regulatory requirement” come with a strange level of holiness. Thou shalt not question them as they are coming from “up high”. This view misses two points:- You most likely are not getting first-hand information, so whoever is telling you “the requirement” probably uses the strictest possible interpretation of it. (Sometimes out of risk aversion, sometimes out of laziness). - Regulators are people, too - it’s possible to talk to them, and if what you are trying to do is useful or new enough, they will help you find a way. Having done this is one of my fondest stories.There are, of course, less serious cases, too - I’ve seen so many requirements that were handed over generations of managers and no one knew what they were for, but it was easier to keep them there.
2. Delete parts of the process or product you are building. Since Steve Jobs it’s a cliche that building anything is about what you say no to, but it’s nevertheless true. Saying yes is easy, but it will slow you down or sink you in the worst case. You can add a few things back later if you really need them. But first, get rid of all the things that got added “just in case”. If you don’t add things back in 10% of the time, you haven’t deleted enough.
3. Simplify and optimise. The most common error you can make is optimizing something that should not exist.
4. Go faster. Everything can be done faster. But again, there is no sense in making something faster that should not exist. (Read this again when someone is trying to sell you their fancy RPA product.)
5. Automate. Even smart people try to automate first. Elon did it with the first Tesla factory. Steve Jobs did it with the factory for the original Macintosh.

I like the last part of The Algorithm best: The only rules are the ones dictated by the laws of physics. Everything else is a recommendation.