[swyx]

> 5. Get in the habit of Fermi estimation, looking up key quantities, and using upper and lower bounds.

i cannot agree with this enough. there are so many people in my life (work and nonwork) that make vague statements (whether about proportions, predictions, or population) that are more directional than precise. this leads you to commit to actions than you then later backrationalize.

forcing yourself to make fermi estimates and upper/lower bounds helps you understand when your assumptions are wrong and change on a dime, avoiding sunk cost/commitment fallacy. but people are very uncomfortable when you make them put confidence intervals around their predictions.

[dools]

It’s funny I never knew there was a name for this but it is definitely something I got in the habit of when studying engineering and I use it all the time by creating little spreadsheet models. You don’t have to be that accurate, but you can create models with somewhat reasonable values and then fiddle with the outcomes and it just tells you if something is even remotely possible or not.

Like hey I might be able to increase my revenue by 10% next month but I definitely can’t increase it by 1000%. It’s surprisingly powerful.