Great article and it is actually making me rethink parts of the design of my own start-up, which would have lots of network value, but it needs some sort of personal value as well. Now that I have a language to wrap around the problem it will be easier.
In the beginning all sites don't have enough data to satisfy any users. That is part of the process of building the site and creating a user base, but in the end if you aren't using it, then why would anyone else?
I guess there is some correlation with scratching your own itch first. It's never bad advice to tell someone to build something for your own enjoyment before anything else. So I guess that applies here as well.
The fact that it isn't obvious for somebody is strange. After all this is how old software works in 99% of the times: you have to write something of useful for individuals alone.
Web applications are now much more collaborative, and things like digg/reddit/... are useless without a community, but when possible it's much better to actually use the fact that individuals obtain advantages using the system in order to create part of the collaborative value.
As a side note, my feeling is that most of the time web applications that are useful for single individuals tend to be more solid in some way: to compete with gmail you have to design a great web mail, since gmail is already pretty good, and this is hard.
To compete with a site where the value is mostly in the community a competitor may attach you in different ways and it is much more a matter of fashion that technology.
I'd love to see a lot more written about this and more sites analysed.
I'm guessing Reddit and Digg took off not such much because they are personally useful without other people there but because news is something that is constant, compelling, frequent, neverending and you could literally hang out there all day finding good information.
Where as sites that are attempting to do the cold start and involve something like reviews and recommendations on restaurants or other services that you only have need to look up occasionally would have a much harder time getting going without being personally useful in any way.
There seems to be a lot of sites attempting that these days. Anyone know some good tricks for overcoming this?
I'd argue that reddit, digg, and wikipedia had enough data to be useful even when they only had 1 user. A single-user reddit is indistinguishable from a blog, but plenty of blogs are useful.
Another way to start your social network is to gradually borrow someone else's. Youtube was immediately useful because links to Youtube videos could be emailed to friends and/or embedded in blogs and other social networks. The site played well with others, which eventually encouraged all blog readers, emailers, and MySpace users to start visiting Youtube directly.
You left out the super-secret invitation-only beta site method, where your first users don't mind the bare walls and the empty pages because they're so excited about being in on the ground floor of a New World-Changing Venture. After your first 10 users put in some data, you let the next 10 in. After you've got enough data that the site seems useful you start marketing in earnest.
A variation is the Craigslist method: use geography to your advantage. Start by building a Craigslist for your office building, then expand to your neighborhood, then your city, and eventually conquer the world. But this might just be a special case of point (1).
While I think this is good advice, there is still the problem that if users don't see much data on your site, then they will assume it's not good and/or trustworthy and leave.
In the beginning all sites don't have enough data to satisfy any users. That is part of the process of building the site and creating a user base, but in the end if you aren't using it, then why would anyone else?