Demographic Markup Language (DML)

Over the last few years I've been becoming more and more annoyed with how targeted web advertising is choosing to become. Between Facebook data mining and Google ad targeting, it seems like every ad on their websites I visit believes I want to get an IT degree from the University of Phoenix and participate in every Tough Mudder event happening on the West Coast simply because I've indicated I'm back in school and enjoy running events.

I laughed and was intrigued when Amber Case recently tweeted her blog post about creating a demographic markup language (DML). Essentially, such a markup language would allow one to define one's demographic information for all web services. For websites, the more data you have on your user, the better you can targets ads, and these ads bring more advertisers and, most importantly, money. So now it seems everyone collects some small bit of information so ads can be targeted. Amber notes, however, how sad it is how significantly one's user experience can change simply by changing basic information since most websites currently have only very small data points, like age or sex, with which to target their advertising. She herself set her gender to male and birthdate to 1973 so she could actually get advertising she likes. By defining a DML, the user would have a better say in the types of advertising they want to see. This is a win win for user and advertiser.

This would seem to pull the carpet out from under social media giants like Facebook. I've always felt that Facebook benefited more than its users simply because of all the data they have on their users. By giving the user more choice in the type of advertising they will see (since they have to), all websites will have equal access to user data and create a richer and more customized web user experience. This could be powerful for many different reasons and start to give identity to the faceless users of web services.