Chris Wiggins describes the practice of data science as the pain that you feel when you find yourself awash in data. For most of his career, Wiggins has worked in academia; he has spent the last decade studying molecular biology at Columbia University, where a data-driven approach has become integral to his research. Despite his breadth of expertise in the field of data science, Wiggins sought a new challenge when he took a semester sabbatical in the fall of 2013. One of his friends advised him, “Go to The New York Times, it will be weird.” He took the advice, and as a result, Chris Wiggins found himself as The New York Times’ first Chief Data Scientist.
Technology and journalism have a tenuous relationship, but it has become clear in the past few years that traditional long-standing media conglomerates have to embrace the digital revolution. It is an understatement to say that the internet has drastically changed journalism, as daily coverage has morphed into a 24-hour digital echo-chamber. With the emergence of branded content, viral media, and the almighty ad-click rate, it is impossible to draw the line between real news, and shareable content. The challenge being: with so much data available, how do you improve your business model without letting it affect the craft of journalism?
Wiggins’ entrance to The New York Times came at a particularly rocky period for the publication A few months after he was brought on, a leaked internal report showed how the Times’ business model had failed to reflect the changing tides of print media. Their social media efforts were scattered at best, and the back end of the Times’ website was revealed to be incredibly inefficient. Many of the news editors were described as being “unfamiliar with the web.” But the most troubling aspect of the report was an undertone suggesting content consumption trends could render The New York Times’ hallowed journalistic practices financially unfeasible.
Most media companies that heavily rely on data science collect information from consumers to decide which types of content to publish. When data scientists at a viral content producer recognize postings of cat pictures are generating more views, and as a result more ad-dollars, than other posts, the editorial teams are informed to produce more media that follows. The result is a media landscape where share-ability trumps integrity, and catchy headlines become more profitable than careful reporting.
Chris Wiggins has a different approach to how The New York Times should use data science. Media websites are no longer just a company’s online presence, they are a microscope into a reader base, and Wiggins believes this is the key to securing journalism’s future. Even though The New York Times’ data science team knows how articles are performing, they have made a conscious decision to not let this affect editorial planning. Whereas the internet’s countless content churns use data science to calculate the types of content generated, Wiggins’ mentality is that data science should be applied from a financial standpoint to support business operations, letting editorial teams function independently. Wiggins is adamant that there is a need for a new sustainable financial model to support the traditional craft of journalism, and although its feasibility is unclear, he believes data science can be the basis of this new model.
In the past year, a major overhaul has changed how The New York Times sees itself. With all the data collected via the Times’ website, Wiggins and his team can take the guesswork out of financial decisions that have plagued media companies. Need to know which website layout keeps readers around longer? The Times is now using A/B testing to determine which features readers prefer. How many newspapers need to go on each newsstand around the world? The Times now taps into previously recorded sales, and other measurable statistics, instead of using industry rules of thumb. What are the indicators leading to subscription cancellation? Wiggins’ team can now spot larger trends earlier to keep a robust subscription base.
At the core of Wiggins’ work is the idea that traditional journalistic practices are essential to our country’s cultural fabric. He believes data science can be used to support and sustain this tradition, as opposed to fundamentally changing the way it operates. Wiggins argues that the internet’s affect on journalism has forced news outlets to think like startups, as opposed to traditional media conglomerates. Two of the indicators that a startup will be successful are whether or not its business model is scalable and repeatable. Is this the case for The New York Times? “We don’t know yet.” Wiggins responded. It remains to be seen. But Wiggins seems confident that applied data science will modernize the Times’ business model, while keeping traditional journalistic practices alive and well. And judging by the drastic turnaround The New York Times has seen on its digital side in the past year, his plan might be working.
Article by Jack Lowery