He's right. The most important thing we tell each YC batch, immediately after interviews, is start now. Startups should be building stuff and talking to users, and not much else.
having done a startup with no "network" at all, and one knowing of and taking part in the tech startup community, I do think there is massive amount of value added when you are a part of "the scene"
You gain a network who will push your startup, introduce you to customers, investors, partners, media, you get some knowledge shared from people in the same situation, tools and methods they used and sometimes its just comforting being part of a group of people in the same situation who you can all relate with.
obviously a lot of this comes as part of the package with YC, imo one of its biggest attractions.
(and obviously this definitely cannot be done in place of building a product / talking to customers)
Introducing people is easy for us, but we need some ammunition to make introductions effective. We need to be able to say not just please meet x, but please meet x, who have built an amazing y.
At the risk of being pedantic, and because I know a large percentage of folks do not understand what you are saying, by "amazing" do you mean something with a lot of users and proof of traction? Not something that looks cool, is an incredible program, or most programmers would think is awesome? Because the average user of HN's opinion of "awesome" and yours might be very far apart.
When I say to investors that something is good, I mean good from their point of view. Which means that users do or will love it, and the market is big.
I think this is something that many people (Industries?) mistake about networking.
Your version is better then the alternative, and both are leagues in advance of "Hi, I'm X, I might want Y from you in the future". Leagues because it's an approach that's nonsensical and belongs in the past.
there is massive amount of value added when you are a part of "the scene"
I won't spam you with a link to our new startup that lets do you exactly that (just check my profile), but I can't stress enough how important the network is. The number one mistake I have seen technical founders (like myself) make is to code one more line of code, rather than talk to a real user or an influencer. There is always one more line of code to write.
There's a multiplier effect at work, though. If your product sucks, no amount of networking is going to save you, because everybody you shop it to will just say "You're f___ing crazy if you think anyone will ever buy this." While if you build something awesome but nobody knows about it, well, nobody knows about it.
The sweet spot is to have something that's good enough that users will sit up and say "Hey, that's kinda useful", and then get it in front of lots of users.
Building up my network is exactly what I'm focusing on right now, because when I came up with the idea, I quickly realized nobody I knew could directly help me and what I want to do is going to require a solid team. I figure I'll spend another few weeks or so on hardcore networking (YC deadline, coincidence?) and then shift over to really building the product.
One of the nice things for me is that entrepreneurs/engineers also count as potential customers/users so I'm still advancing my cause this way. By not being shy about what I'm trying to do, I've gotten a lot of genuine interest and useful feedback.
I guess what I'm trying to say is that network effects are a major component of my momentum right now and that even though a focus on developing the product is good, building a good team quickly and finding the right support are worth the upfront investment.
Obviously, building product is the sine qua non of a startup. But, there's a huge difference for companies that can build a great product, and then access the YC brand and network, and those that don't have access to that brand and network.
There's a number of companies that build a great product, but then they need access to a network for advice and funding. If you're going to get funded via Angel List, for instance, you have to have some form of social proof.
