I absolutely agree with the writer's point about over-engineering. In start-up / small business land, optimizations are worthless because the main problem isn't

I absolutely agree with the writer's point about over-engineering. In start-up / small business land, optimizations are worthless because the main problem isn't that "oh, I have a poorly performing inefficient system", and is instead "I don't have a system and need to build one." So to a start up founder, timing correctly when to do something is far more important than doing that thing amazingly well (passable is enough). In other words, technical debt is not a bad thing (just humbly apologize for it later). And regarding building a startup by yourself, I'd like to expand that to never build a network-related startup by yourself. That is, if... value = k2 * users ^ 2 + k1 * users + k0 never build a product where k0 = 0 or close to it by yourself. This means, for example, if you're by yourself, build an indie game where you jump on platforms, but never attempt "facebook for dogs". This is because you can't get customers to pay $ for no value, so building social networks without an extremely strong initial user base in your pocket (which, unless you're a celebrity, you don't have) is the tech equivalent of buying the lottery. In a lot of ways, social network features on an app is like legal - if you need to worry about it, you've already gotten to the point when you can delegate an employee to do it.