> I've never seen a start-up fail on account of technical debt.

That's because noone ever says "XYZ startup failed because their code quality wasn't good enough." It's usually something closer to "Their customers didn't renew contracts, they had to slash headcount, and went into a death spiral from there."

If you press for details, the deeper reason might be "they couldn't turn around features fast enough" or "the app was slow or their servers went down too much" which is a symptom of too much technical debt (either directly, or because the better engineers on the team choose to leave as a consequence of the technical debt).

I agree that early startups don't fail because of technical debt, but I'd bet that many mid-size startups do fail because of it.

Ok but like middle ground exists here. There is a huge difference between thinking carefully about what you need to do right and planning what bits are going to end up as technical debt, and how you can fix it later - and what actually ends up happening in my experience where everything is shit because we're too busy being a s t a r t u p to practice engineering.