I'm a math, machine learning and information retrieval geek. I've never applied it to a members section of an adult site, but I've been thinking of doing it for a while. This may sound a bit far fetched, but let me try to explain what I am thinking
Each visitor is interested in a specific set of content and attempting to sell him/her on something else won't work. The trick is figuring out what kind of content that specific user is looking for. An easy way of doing that would be using the visitor history of all the users of the site to see which types of content are viewed together and automatically make "clusters" that describe each of those content types. Then you can assign a new visitor to one of those clusters, based on what content that specific user is looking at (and has looked a previously), and show them only affiliate content that is related to his/her interests. There are of course many other ways of recommending content to users, but that would be an easy way to start.
I'm not really sure what my point is.. maybe I should develop a product and market it to site owners
Anyway, now you know that there are options for optimizing affiliate sales based on the interests of each user and I think this industry could benefit greatly from using algorithms to maximize profit.