This data is not provided directly by users. Although they consent to the transmission of information, it is the companies where they register that they end up selling the same information to other companies. It is thus possible to create very detailed profiles of each user as they give different types of data to each platform and service, revealing different parts of their identity and customs as they broaden their digital footprint. For this reason, Johnny Ryan, who is responsible for the policies for using the Brave browser, believes that in the next US election, "it is inevitable that every voter is profiled based on what he has read, watched and heard online over the past few years.
But the set of information that advertisers have access to, does not stick to what they know clearly and unequivocally. Suppliers claim that, based on the data, they can predict behaviors and tastes. This means that even without having their voter registration, companies can determine who is most likely to vote in the upcoming elections or where they are going next holiday next summer. This is possible through behavioral comparison.
By comparing the pattern of use and navigation of a given user with multiple data groups, data providers can predict, for example, whether a user will actually buy a car. To do this, just understand how the users who bought cars recently, behaved online. By identifying patterns and users that begin to fit into them, you can see where the greatest potential for business conversion lies.
This causes the user to be automatically placed in a batch of users to whom a car manufacturer will be more interested in showing a promotion for a new car model, for example. It is important, however, to understand that the forecasts are not factual, but bets, which despite having a high percentage of success, still present some margin of error.
The lack of privacy entailed by this business model is problematic, but when brands are not the only entities that can sift through this information, the risk is even greater. In 2017, for example, a group of German researchers managed to link a data set, supposedly anonymous, to the individuals who generated them. The results of this experiment were revealing of the risk potential inherent in information leaks. By linking "die" with "user", the investigators found out what kind of medication a German politician was taking and what kind of pornography a judge preferred.
Targeted ads are potentially better for brands, given their potential efficiency, as well as for the public, which thus crosses with information more suited to their preferences and habits. However, the data collection that feeds this targeting may be processing at an overly invasive level, which touches on very personal dimensions of users.
"Do businesses really need to monitor all the websites people visit, all the apps they use on all platforms, and keep those data for long periods of time?" Asks Kaltheuner. "Is this all just to show relevant ads?".
The ad companies register all their use of the internet, through monitoring systems and cookies that maintain an updated profile of their preferences. Then the data is shared with the advertisers, who auction the advertising spaces available online. Brands that put more money "on the table" get the right to advertise in space at auction.
Shoshana Zuboff, a professor at Harvard University and author of The Age of Surveillance Capitalism, believes that this exploration is similar to what businesses do when they turn them into commodities. It is a way of fueling a "surveillance economy".
"This is all about advertising, but all the negative implications that the system entails have nothing to do with ads, and that's the price we pay just to see relevant ads," Kaltheuner said.