This isn't 1950. Half of the population is not crowded around a TV at night watching three shows. For a long time now, traditional TV networks have been struggling. This is no blip. More people than ever are cutting the cord. Traditional media outlets are scared—and they should be.
Against that backdrop, TV networks are looking for answers. They're not blind. They see that Netflix has used what we call big data incredibly well and they're trying to mimic its success.
Against this backdrop, it's likely that networks are increasingly looking at their own internal data, Nielsen data, and even third-party metadata to inform their choices. Take ABC's new show Selfie. The show's title cannot be an accident. It seems geared to the very audience that traditional networks are chasing (read: Millennials).
Are TV networks using data to predict what viewers will like?
Networks are trying to do this. It's important to note that, by itself, data can only do so much. "Data" doesn't create dialogue, characters and plots. For instance, Netflix, used its own internal data to market House of Cards differently to its customers. Fight Club got heavy helpings of the David Fincher angle in House of Cards, as he directed both the movie and the first two episodes of the series.
Make no mistake, though: Netflix could not predict that the show would be successful. The company used data to make better bets, but they were still bets. As I write in my last two books, by now means does big data eliminate uncertainty.
Netflix bet big on resurrecting the cult classic Arrested Development. As I wrote in Wired, the company knows which viewers respond to different colors. I doubt that its choice of color in hit Orange Is the New Black is entirely a coincidence. Again, though, data is just helping people find the show. By itself, data isn't predicting what people will like. We're not there yet.
Can networks respond to data in real time?
The short answer is no.
Networks can certainly adjust to ratings, but just not in real-time. Unlike tweets, YouTube videos, and blog posts, traditional television shows are scripted, filmed, and planned months in advance. They are not streamed live. Perhaps individual scenes can be cut or altered when the media leaks a certain scene and the public reacts poorly. Exhibit A: Even before it aired, Fox came under fire for an ill-advised joke in the much-publicized crossover episode between The Simpsons and Family Guy.
There are so many uses of "data" that it's impossible to give a quick list. There will always be surprise hits, things that data could never have predicted. The descriptive power of data will always trump its predictive power. Put differently, it will always be better able to explain phenomenon retroactively, a point made in many books like The Black Swan.
Simon Says: Welcome to Data-Based TV
That number is only going to increase in the years ahead. Big data has become too big to ignore. The success of Netflix and the challenges faced by networks (mentioned above) mean that traditionalists are finally waking up. This is tantamount to what has happened in baseball. As Michael Lewis writes in MoneyBall, Billy Beane of the Oakland A's used to be viewed as a iconoclast. Now, every major league baseball team sports an analytics staff. I strongly suspect that the same will happen in network television.
Feedback
What say you?
