The thing we look for the most is a small number of users that are super, super engaged with the product.

Most new social networks we see point to their top-level growth of new users. These usually eventually fade.

What you want instead is something that starts with a small number of users that use it many times every day. Snapchat is a good recent example of this phenomenon.

Same with

- pinterest: some collection of moms in middle america

- twitter: SF Tech & SXSW

- facebook: ivy league colleges -> colleges -> high schools -> the world

- snapchat: so-cal teenagers

I recommend validating your market with a newsletter or some other channel that already has social engineering baked in. In the case of Product Hunt & Mattermark, this allowed them to demonstrate value with curation of highly valuable content for a particular group of people. Once you have validated it, creating the application seeded with those users will ensure you have the well in place. Spending time building the app without having the ability to iterate first on the differentiating aspect might be a waste or at least an inefficient use of early resources.