Data-driven journalism is the craze du jour in media circles but, as recent raspberries blown at Nate Silver suggest, even some of the craft’s alleged experts appear unsure how to go about it. And they’re not the only ones trying to make sense of data: it turns out that even media giants like Turner Broadcasting System, whose properties include CNN and the Cartoon Network, are facing data dilemmas of their own.
For a company like Turner, the challenge lies in collecting and making sense of an unprecedented flood of new audience and advertiser information.
“It’s as if the physics of engagement are shifting and we’re trying to solve it at the same time,” said Colin Coleman, an analytics executive at Turner, who suggested the task was more complex than those he faced while working for NASA.
Coleman, speaking at Gigaom’s Structure Data event in New York, said the complexity of “building the house of data” lies in the fact that both the industry and its business model are in flux.
“We can’t broadcast anymore,” said Coleman, explaining that Turner must navigate a plethora of devices and technologies just to reach and keep track of its audience in the first place. Meanwhile, advertisers are demanding that Turner employ tools that can identify and target people in real time.
Coleman cited an example familiar to any digital media company: how to decide the relative value of a visitor who arrives to a website via a Google search versus one who comes in via a social channel like Facebook. As it turns out, one type of readers may spend a lot more time on the site than the other — meaning that a single “social” visitor may be more valuable than two “search” visitors. The calculation becomes even more complex since each type of visitor’s digital fingerprint may come with more or less of the personal data advertisers are seeking.
Coleman said that while Turner is rapidly figuring out how to parse all these factors, the company still has more data than it can wrangle. He conceded that, from a technical standpoint, the company is for now “Hadumping” (a reference to Turner’s use of Hadoop, a framework for collecting and processing large amounts of data).
He added that the task of organizing the data is more difficult because of its diverse audience — from NBA fans to young children — and ensuring that Turner complies with various privacy rules. Unlike NASA, he said, the scientific challenges come with a huge amount of compliance and governance issues — tasks made all the harder since big media companies are still organized in silos that predate the digital age.
Photo by Jakub Mosur