Chris Anderson thinks that popular content should be free (and ad-supported) while niche content should be charged for. This is obviously a reaction to the Wall Street Journal move towards payments but also a reaction to the past attempts by, say, the New York Times to charge for premium stuff like opinion pieces. This model would seem to make sense if it is the case that popular stuff is very price elastic unlike niche stuff. That I can believe but the issue is: how do you work out what is popular and what is niche? Or to be more specific, how do you work it out before the fact as it is easy to tell after the fact?
One option is to have a subscription model and then rely on intra-site traffic measures to quickly re-assign content as popular or niche. The popular content would then become freely available although with a lag. The niche or, actually, the unpopular would stay that way. Subscribers get a first look as well as a complete look.
Another option is to use micropayments but with an out: if an item becomes popular, it becomes free (and you get reimbursed). So I might visit a site, read an article but if I refer it to enough people who then visit the site, it becomes free for me. This has the advantage of providing an incentive to drive traffic for popular items and also to absolve some of the ‘free riding’ that might come from waiting for popularity. For people who write blogs, this provides an incentive to link to interesting articles so as to save on payments.
I’m not sure these things will work as for the most part I could not tell whether something will be popular or not and so would realise that I am carrying a financial risk in reading. But it is fun to think about whether there are optimal pricing mechanisms here.