Everyone's familiar with how businesses use our personal information to deliver targeted advertising online. But new developments in predictive analysis goes far beyond that, allowing companies to use thousands of points of data to make startlingly specific assumptions about customers.
Network analysis, also known as predictive analysis, uses data about your interactions with others in a social network to make assumptions and predictions about your behavior. Basically, they read your mind. The Economist shows how powerful this can be for companies looking to sell products: Telecom providers, for example, want to identify "Influencers" who might persuade others to switch to their services. How do they spot these influencers?:
People at the top of the office or social pecking order often receive quick callbacks, do not worry about calling other people late at night and tend to get more calls at times when social events are most often organised, such as Friday afternoons. Influential customers also reveal their clout by making long calls, while the calls they receive are generally short.
This kind of stuff is quickly becoming corporate dogma. According to the Economist, some financial firms are using network analysis to determine risky borrowers by looking at their social ties; startups like Socialcast turn the analysis on companies themselves, helping them see where employees fit in the corporate social graph based on their digital interactions with coworkers; and companies could analyse all this data using one of IBM's "Next Best Action" programs, which tells retailers how they should respond to a customer purchase.
This is of course a boon to businesses who would like nothing better than to stick an electrode on your brain to suck up your every thought. But would you believe there's a troubling side to network analysis, too? For one thing, Internet retailers already screw over existing customers in favor of attracting new ones with sweet deals and better service. Imagine if they started using network analysis to give 50% discounts to the well-connected teenager they thought would bring in all their friends, while offering 100-day shipping via beat-up Dodge Caravan to the poor Appalachian farmer who just wants cheap diapers for her 18 kids.
And it's the promise of these kinds of technologies which is in part driving Facebook and Google's relentless privacy rollbacks. The more data you give up, the more valuable these services become to advertisers. Pretty soon, "Would you like fries with that?" will be "Based on the number of links you have with key influencers who have purchased Atkins Diet books in the past twenty months, you do not want fries with this."