startupquote
startupquote:

Be suspicious of any work that is not building product or getting customers. It’s easy to get sucked into an infrastructure rewrite death spiral.
- Sam Altman

So, I like this quote but I think it needs a bit of clarification when it comes to getting yourself into that “infrastructure rewrite death spiral”.
Rewriting infrastructure is not a BAD thing. Fixing fundamentally flawed ideas you had at the beginning of the startup to “get it done” is not bad, but there needs to be a balance between managing your technical debt and the product.
The product cannot live and thrive if the infrastructure is pulling it down into a bog of mediocrity and the infrastructure will not live and thrive if there are no customers/product.
Be suspicious, absolutely, but understand that it’s going to happen. You can’t write 100% perfect code or architect a 100% perfect system upfront, with little to no information about users or the problems you will face.
Once you’ve hit the point where all you’re doing is iterating on your infrastructure without touching the product, you’re dead in the water anyway. The early startup thrives on quickly making decisions and moving through product iterations (as well as infrastructure iterations) as quickly as humanly possible to pivot against all problems.
Listen to user needs, iterate, pay the tech debt often and in bite sized chunks, and everything will work out fine.

startupquote:

Be suspicious of any work that is not building product or getting customers. It’s easy to get sucked into an infrastructure rewrite death spiral.

- Sam Altman

So, I like this quote but I think it needs a bit of clarification when it comes to getting yourself into that “infrastructure rewrite death spiral”.

Rewriting infrastructure is not a BAD thing. Fixing fundamentally flawed ideas you had at the beginning of the startup to “get it done” is not bad, but there needs to be a balance between managing your technical debt and the product.

The product cannot live and thrive if the infrastructure is pulling it down into a bog of mediocrity and the infrastructure will not live and thrive if there are no customers/product.

Be suspicious, absolutely, but understand that it’s going to happen. You can’t write 100% perfect code or architect a 100% perfect system upfront, with little to no information about users or the problems you will face.

Once you’ve hit the point where all you’re doing is iterating on your infrastructure without touching the product, you’re dead in the water anyway. The early startup thrives on quickly making decisions and moving through product iterations (as well as infrastructure iterations) as quickly as humanly possible to pivot against all problems.

Listen to user needs, iterate, pay the tech debt often and in bite sized chunks, and everything will work out fine.