I have an interesting problem, I am trying to estimate populations (ages mostly) of users who visit my website - I have IP->ZipCode->Census demographic information (which gives me an age distribution per zip). Now clearly this site, which is geared towards young people, is a much smaller subset than the entire population of any given zip code.
More specifically, I am interested in how these populations change per blog post / sub-page.
It would be great if someone could point me in the right direction, or specify which techniques might be helpful here.
Thanks!