Notional Framework Breaks the Problem Up
This framework is an extension of the ideas presented in Monetization for Web2010.
Social media provides the "real estate" for communications to happen. Whether social media is decentralized (ostatus, P2), or integrated by API and more centralized (twitter, facebook) the conversations are happening because people are connected. Without the backbone of social media this monetization would not be possible so they must be aportioned a piece of the revenue pie (20-30%?).
Web hosters (bloggers, companies, individuals) can engage in conversations on their own sites while sharing windows into the social media world. Folks that host shared media on their site empower and evangelize the social media brand (as I do with friendfeed with window real estate and twitter by content). By hosting their own content and seemlessly integrating their favorite social media "hang outs" individuals shoud see a cut of ad revenue (40-50%?).
Existing advertising networks (Google, Microsoft Ads etc.) have developed leading edge technology (algorithms) to match semantic meaning (meta data) from text to drive appropriate ads. In addition they provide a single stop for businesses to advertise to (business advertising aggregators). We can leverage their technology moving forward. (20-30%?).
But I'd like to see semantic algorithm businesses decoupled from advertising aggregators. That would comprise a four part business model allowing for greater competition and innovation within each space (with overlap). For example friendfeed could team up with wordpress to create personalized user profiles to better serve ad algorithms. Google could team up with web hosters to integrate Wave real estate directly into their web pages (like friendfeed has done).
Edit: I added input from the meta data extraction algorithms to the front end sites, as well as user like/dislike information back to the social media platform and advertising aggregator.