The Shadowy World of Data

I was recently interviewed by Matt Kapko from CIO Magazine for a piece called Inside the Shadowy World of Data Brokers.

Matt had shared with me a 60 Minutes story on " The Data Brokers Selling Your Personal Information ." It paints a pretty scary and sinister picture. Morally, I am very strongly opposed to abuses of privacy. I really find it unsettling that Axciom will sell you information on things like how much I paid for my mortgage—and according to 60 Minutes—what kind of psychological disabilities I might be suffering from (as if my writing hasn't exposed them to you already).

But, I felt there was some flaws in how 60 Minutes presented the data on Big Data. You see, not all data is created equal.

On one hand, you have the same kind of data that data brokers like Axciom have been selling to direct marketers and telemarketers for many years. This is the really personal stuff with financials, your phone numbers, your kids names and other stuff you might not feel comfortable with strangers knowing. This is yucky, but it's nothing new.

On the other hand, there is the anonymous cookie data a zillion companies are using to track every link you click on, every term you search on and in the case of mobile phones, every place you go. "Holy crap!" you might say, but the reality is that this data is generally just used in aggregate to identify patterns and serve you ads that hopefully are more relevant to you. I'd classify serving you ads based on aggregate patterns in data as nervy, but certainly less sinister than the first scenario above.

Now, however, they are matching that behavioral cookie data to the personal profile data that can be purchased on you. So not only do they know the amount of your mortgage, but what kind of websites you visit or places you bring your phone.

I've pasted part of the article below, but wanted to give it that context. You can read the whole story (and watch the 60 Minutes piece) here .

"[Data has] become an essential part of the marketing mix," says Adam Kleinberg, CEO of Traction, an advertising and interactive agency in San Francisco. Data brokers are "becoming increasingly important because the way digital media is being purchased is moving toward the robots. Programmatic advertising and programmatic media buying is using tools that automate the process," he says. "You enhance the targeting efficiency by leveraging that data. It's just gotten to the point in the past few years where 30 to 40 percent of media is purchased that way."



These profiles are directional and optimized behaviorally, Kleinberg says. The cookies that follow us around the Internet are being used to index us based on behaviors such as what we search, visit, click on or buy. "If you actually saw your data you'd think 'wow, these people don't know me at all,'" he says.



"The power of the data in certain circumstances is in the massive quantity and patterning that is possible. When you're collecting across billions of data points, regardless of its accuracy, there's going to be groups of individuals behaving the same way," Kleinberg adds.

"There is sensitive data that is collected and sold on you... What's new is this big data that is being collected and cross referenced with those things," he says. "The reality is that most of this big data is simply being used anonymously to better target you with an ad."

While he freely admits "the ability to look at that individual data is a little scary," he adds that "anyone who's buying digital media today is buying data ..."

Sharing the view of the industry at large, Kleinberg says he thinks the responsibility should come from within because regulators don't have a deep understanding. "I think that the industry organizations are actually taking it very seriously and putting together standards that accommodate reasonable privacy restrictions like allowing people to opt out," he says.



"I think consumers care less than we think in the moment. They care in the abstract sense," Kleinberg says. "I can't tell you of an example where data has been abused."

To embolden the case for self-regulation, the industry needs to do more to explain what data means, Kleinberg adds. "The terms data and big data get lumped together as this big sinister beast and a lot of it is not innocuous ... it's anonymized by obscurity," he says. "We should not rush to judge all of it without understanding that nuance."