Despite all the talk about how better advertising-targeting will improve the economics of online publishers, advertising networks and experience of users, reality that it still much talk (some of it by me) and little to show for. Some datapoints:
* Social network Tagged allegedly makes about $600 000 on a billion pageviews per month from search and banner networks. That is about $0.6 CPM. It makes Tagged profitable, but not many web sites have a billion page views per month. Unless your site offers significantly more targeted traffic than a general social network, you are unlikely to make much money from AdSense or advertising networks. You should figure out a direct sales/branded advertising strategy for your site.
* Google pointed out in their latest earning call that they have problems with (improving) monetization of social network inventory.
* MySpace still says its Hyper-Targeting program is improving the rates the site can charge. "What is an issue - making sure that MySpace's hyper-targeting system, designed to match up advertisers with users based on profile data, continues to work. News Corp. continues to boast that advertisers who use the system have seen click-through rates increase up to 300%; yesterday they also said that they've been able to raise rates for those ads by up to 50%, and that advertisers who use the system spend twice as much on the site."
If there is a point with this post, I think it is this: It is hard to make much money on a social site if the majority of your sales is handled by an advertising network (even if that network is Google). There is quite a bit of work being done in improving the relevancy of advertisements shown to individul users and thus, in the end, the revenue for the social network. However, the code hasn't been cracked and shown to scale for advertisers. Today's and future's realities are two very different things.