It has its place, but just as useful at times is getting together with a group of peers you trust and exchanging notes. Having a few beers and talking about what challenges you're having, what breakthroughs you've made...
I think the tradtional idea of networking is overemphasized among a lot of startups (and sometimes dangerously so) because it's an abstraction away from the work that keeps them alive. You've got to make sure you keep a watchful eye on your schedule and don't confuse attending X and Y event for Z productivity.
Not sure I'm saying "my kind of networking meets X times per month", but I will say this: we organized monthly meetups for the professionals in our market (computer/network security), made it as all-inclusive as we could, and it's been far more valuable to us than the vanity networking we did early on.
Isn't this one of the major recurring criticisms of valley culture? That all those people do is go to parties with the same people every night?
I'm with you on that.
SXSW trips not so much.
I don't think Wozniak did much networking. I don't know what would have happened if Jobs got mad at him because he didn't.
I'm not sure I'd question the commitment of people who were actively building the product either. It sounds like this startup really had some forward momentum with people getting stuff done. I'm not sure the authors expectations were reasonable, and I think he had to work fairly hard to kill this startup.
I agree with the second part of your comment.
On the other hand, some folks take being pragmatic to an extreme, as in no order/uniformity/quality control/structure/thought into design/effort to identify possible future issues and a general just get whatever done in a hurry and push it out the door attitude . . . which is all just as bad as too much formally define roles/establish this and that and the other/talk about and plan stuff to death/posture around with this and that and the other . . .
If you have good people, some funding, a demo and the start of a code base then thats a lot to lose over your expectations on networking and days of the week devoted to the startup. Even if the truth is more networking and fulltime cofounders would have helped.
I certainly may have killed the startup by emphasizing things that I thought were important, but it doesn't make any sense to think I was actively trying to kill it. My point was that I was blindsided by what happened in the end, so now I've gone back to think about how we got there and, you know, these are some of the conclusions I've drawn and of course they could still be wrong.
On the one hand you could feel in retrospect that the conflict points were important enough that the startup could never have succeeding without the direction you were pushing - so you were lucky it ended swiftly. On the other hand if you think your startup had a good chance of success even if you hadn't won those battles, then you may want to look at your conflict resolution.
After spending a few months thinking "okay, so how did it get to that point, and I didn't even realize it" I wrote my reflections. I still think the startup has a good chance of success, but not with the dynamics that were at play then.
It totally depends on the product, but from the YC PoV, you have 3 months to build something that dazzles investors. Generally, YC discourages hunting for networking/bizdev/funding deals because you have EXACTLY ONE SHOT to impress a room full of the most important early stage investors on the planet.
You can always network after you've got a product, but if you can't do both at once, product comes first.
Edit: What I mean by this is networking at the expense of product/technology building.
As for allocating resources to networking before there is a working product, check out www.path101.com. I met these guys when they did an open brainstorming session with nothing but a preliminary pitch deck and not a line of code. Being open from very early on enabled them to find angel investors and their first employees, and when it came time to doing an alpha launch they had tons of people that had been eager to try the thing out for months.
What happens when it turns out that the bullet points with features cannot be implemented, or are simply not the best solution for the users?
I find pg's thoughts on the subject of how problems may end up being redefined as they are being worked on to be very wise advice.