At what point does “move fast” start becoming expensive? - I will not promote

​ I’ve noticed something with startups. Early on, cutting corners makes complete sense. You don't have time to build everything properly, and honestly, you shouldn't. But some of those shortcuts quietly become permanent. You skip proper testing because you're trying to get the MVP out. Then you keep doing it because you're trying to hit the next milestone. Then suddenly the team is spending more time fixing old stuff than building new stuff. The weird part is that nothing necessarily breaks immediately. Sometimes the shortcut works for months. So where do you draw the line? Is there a point where a startup should stop saying “we'll fix it later” and start investing in things like testing, documentation, monitoring, processes, etc.? Or is that just something people convince themselves they need too early? Curious how other founders/early-stage teams have handled this.