The Sunday NY Times Magazine recently published a provocative article, "Who Do Online Advertisers Think You Are?", by Jeffrey Rosen, a law professor at George Washington University and the legal affairs editor of The New Republic.
In the article, Rosen describes the rise of online advertising and emergence of data-driven personalized display ads. He focuses on the privacy implications of being tracked by marketers over time — a state of being described by U. Penn Prof. Joseph Turow as the "long click."
Rosen paints a bleak picture. Consider the following quotes from the article:
"The long click raises obvious privacy concerns. But the privacy threats go beyond troublesome ads…computers can link digital profiles with our real identities so precisely that it will soon be hard to claim that the profiles are anonymous in any meaningful sense."
He goes on: "Companies can combine hundreds or thousands of facts about us into what [Professor Paul] Ohm calls 'a database of ruin.'"
And on: "… profiles that define us forever can also be technologies of classification and exclusion," branding some consumers with what privacy advocate Jeffrey Chester calls "a digital scarlet letter."
Data-driven personalization "leaves no possibility for individuality, eccentricity or the possibility of developing tastes and preferences that differ from those of people you superficially resemble."
This data dystopia sounds downright frightening. But are the ominous outcomes Rosen predicts in the article inevitable? Are they even likely?
Rosen himself points out the huge gap that exists between today's anonymous profiles and the omniscient profiles he foresees in the future. He conducts his own experiment, creating a "Republican Jeff " and a "Democratic Jeff " by surfing to different sites on different browsers. When he evaluates the segments compiled by data-aggregator BlueKai, though, neither profile turns out to be particularly accurate.
It's true that many companies, including mine, are working hard to improve the accuracy of data signals. On the other hand, the combination of the enormous and increasing amount of data being generated together with industry codes of conduct ensure that marketers will only ever have a limited view of our activities and identities.
I also take issue with Rosen's implication that data-driven marketing will produce a future in which consumers will be doomed to only those offers and opportunities with which marketers present them, based on their uber-profiles. Such a vision overlooks the much larger context of the digital age. The data-driven Internet revolution has fundamentally empowered individuals in unprecedented ways when it comes to finding what interests them.
For example, these days I'm in the market for running shoes. Rosen's sequencing presumes that I would be bombarded by shoe-related advertising dictated by data-company profiles, relegated thereafter to choosing among the small number of shoe offers presented. Instead, the opposite is occurring. With the enormous variety of information available through the Internet, I am able to do research on running shoes across diverse sources. Based on the interests I express through my research, I may be presented with downstream advertising offers, which I can take or leave.
Far from limiting my choices, data-driven advertising helps underwrite the independent running sites that expand the information and offers available to people like me. As the cost of digital access continues to drop, this cornucopia of digital content is becoming available to literally billions of people around the world.
It's disappointing to read articles and studies regarding data-driven marketing that fall back on the same vague and tired predictions of future harm to consumers — predictions that consistently fail to cite even a single example of a person who has been harmed.
For all the dystopias some writers like to conjure, big data is here to stay. If anything, it's just getting started. In order to assess the real risks and opportunities of the age of big data, we need to focus on known implications, including both actual benefits and demonstrable harms, rather than marinating in narratives founded on predictions of future harms.
Frightening people will always get their attention. But in the age of big data, let's look at the real numbers — of publications, blogs, tweets and other liberating sources of information — as we recommend policies that can either enhance or restrict individuals' Internet-enabled power.