Premature scaling is the #1 killer of startups. But never scaling will also kill you.
At some point you’ve gathered enough evidence and need to start putting resources behind the thing.
We, as a community, spend a lot of time encouraging ourselves to be more careful, more cautious, to gather more evidence. That’s because, historically, the default mistake was to do the opposite & over-commit too soon.
Similarly, I think that working nights & weekends is an awesome way to start doing customer development and answering the big questions about your business. You’ll have a much better starting point when you do take the leap. But that doesn’t abdicate you from the fact that at some point, you’ve got to take the leap, build a proper team for whatever stage you’re at, and get version 1 (or 2 or 3) out the door.
One way to think about premature scaling is as irreversible decisions. For example, it’s very hard to go back to your old burn rate once you’ve hired a VP sales and it’s hard to re-vamp your product focus after a big PR campaign.
It’s irreversible when it makes a permanent dent in the number of months of runway remaining. It’s worth spending as much time as you need to make sure the foundations are solid before deploying that time & money. It’s worth trying to find sneaky tangential experiments which get you the learning without needing to spend all that cheese.
Once you’ve learned all you can at this level and you’re happy with what it shows you: push the button. Pull the lever, flip the switch. Staying trigger-shy once the business needs advancement puts you in a dead end.
 You commit too much of your resources, too soon, to a plan with insufficient evidence.
 Data via the startup genome project.
 I mean organisational scale rather than technical scale — hiring, funding, advertising, etc. Basically anything where the goal is gaining lots of users or customers.