It can be a big competitive advantage to do everything yourself, because you see inefficiencies that fall through the cracks between job descriptions. That's what's left after you've specialized and optimized: all the things about the product that suck because they are nobody's responsibility. Fix them and you not only are doing something that your competitors aren't, but you're doing things they can't do without re-organizing and losing all the accumulated domain knowledge that they have.
The flip side of this is that you need to pick a problem where this is actually an advantage. If you're solving the same problem that your competitors are but they have 100 highly-specialized employees but you have one generalist, you're going to lose. Big. But if you identify the problem that they're not solving because nobody in their organization has the relevant expertise or job description, you can get a solid foothold, then use that to hire the expertise that you were missing in the first place.
> It can be a big competitive advantage to do everything yourself, because you see inefficiencies that fall through the cracks between job descriptions
i agree, but this is begging the question of actually being able to do a good job at any of it. i'm not saying it's impossible, i'm saying most can't.
The point isn't to do a good job of it, the point is to do a better job than the people who are not doing it at all.
This is probably the most counterintuitive part of founding a startup that I've observed so far: if you're doing it right, you'll suck at everything you do. Why? Because if you don't suck at something, it's time to seek funding, hire people, and train them to do it. The founder's job is to seek out problems where everyone else sucks more and yet people still want it, which itself is quite a challenging task.
The flip side of this is that you need to pick a problem where this is actually an advantage. If you're solving the same problem that your competitors are but they have 100 highly-specialized employees but you have one generalist, you're going to lose. Big. But if you identify the problem that they're not solving because nobody in their organization has the relevant expertise or job description, you can get a solid foothold, then use that to hire the expertise that you were missing in the first place.