Web 2.0 is about community. Web 2.0 is about relationships. Web 2.0 is NOT about Ruby on Rails. Web 2.0 is NOT about building a "better" MySpace.
Robert Scoble posted an interesting tirade yesterday about Steve Balmer misinterpreting social networking. Scoble cites Balmer's propensity to respond with the Microsoft insular comment about Web 2.0 "We could build [fill in the blank with Web 2.0 technology] in a matter of [short period of time]" as wildly off base. I couldn't agree more.
Scoble's most compelling argument focuses on Ebay. Ebay isn't successful because of its technology. It's successful because it surpassed the tipping point. Everyone uses it for auctions. Why would I want to use another service to buy or sell something when everyone is on Ebay?
Every time I see a job-board listing needing a Ruby/PHP/whatever developer to "build a Facebook-like" application or a "Monster-like job board", I cringe. So some engaging entrepreneur has orchestrated a way to build a better mouse-trap. So what? You're wasting your time and money.
Of course you could build a better [fill in the blank with Web 2.0 technology] given adequate resources--but it doesn't matter because the social network has already formed around Ebay, Facebook, MySpace, LinkedIn, YouTube, etc. Have you looked at Craig's List lately? It's difficult to look at it's so ugly...and yet, it's the destination for local people-centric, one-time commerce.
So, my advice to you aspiring entrepreneurs: innovate horizontally; not vertically. Come up with the next social or community offering--don't simply repeat what's already there but with a better [interface|design|API|etc]. Can you say Ubid.com? 'Nuf said.