It's a bit trite, but it's a message I want to get across to other startups. It's bit me hard at some of the companies I've worked for long ago - the biggest cu

It's a bit trite, but it's a message I want to get across to other startups. It's bit me hard at some of the companies I've worked for long ago - the biggest culprit was EC2. There are days that they'd simply stop launching instances for us (even though we were nowhere near capacity), or performance between two components would degrade to an unusable point. We were easily spending two full-time engineer's salary every month to AWS for this privilege, with no access to anyone at AWS to immediately figure out what was happening. We ended up redirecting that massive amount of money to hedging our bets and ultimately lowering our costs by minimizing AWS usage. Sad, wasn't what we wanted, but it was the only way that we could make our business work.