3 ms·
A good rule of thumb is: "plan for scale, but don't implement until needed". True, you don't want to engineer yourself into a dead end by designing a system tha
by wolframhempel 8y ago
A good rule of thumb is: "plan for scale, but don't implement until needed". True, you don't want to engineer yourself into a dead end by designing a system that can't be parallelized, but I've also seen so many startup founders (including myself) that design for massive scale that won't arrive for years, but introduces complexities that will slow down development in the exact phase one needs to be the most agile.
- ChikkaChiChi 8y agoThis is good advice. It's all about risk appetite. We should be identifying and communicating both the potential for over-optimization and risk of not being scalable. You can only hope to hit the sweet spot if you have clarity on both being potential outcomes.