The idea is pretty simple:
👉 Deliver one feature per day.
Now, let's describe:
Don't overplan. In fact, plan as little as humanly possible.
Why? So I can use what I build as fast as possible. Since I am trying to use the tool, the most important lever is speed-to-market.
Once I'm using the tool, I can figure out what is worth 1) extending, 2) re-building, or 3) optimizing.
Unlike enterprise software, where mitigating risks is often the name name of the game, I have one risk to mitigate:
👉 wasting my own time.
To do anything more than MVP is overkill and should be avoided.
Instead, I plan as little as possible. I build the plan as I fly it.
Read the manifesto, or, better yet, the principles.You can also go home!