Beyond Your Wildest Streams, Part II: Data Is The New Czar

Studio executives, the great and wicked S. J. Perelman famously observed, “have foreheads only by dint of electrolysis.”

He was speaking of the likes of Louis B. Mayer and Jack Warner, Hollywood despots who harbored nothing but contempt for any picture with artistic or intellectual pretensions. They were unreconstructed balagulas, which is Yiddish for “wagon driver” and never spoken as a compliment. Their fortunes were built upon the lowest common denominator. And though they were scorned by the New York wags, they were also admired for the uncanny insight into popular tastes.

The admiration was misplaced. They weren’t philistine savants; they were just shrewd enough not to try outthinking the bell curve. As Warner himself once admitted: “If I'm right fifty-one percent of the time, I'm ahead of the game.”

Yet the history of Hollywood -- and Madison Avenue -- is the history of taste czars imagining they somehow can divine exactly the right movie, TV show, ad slogan, product based on some combination of instinct, demographics and maybe swelling on their bunions. In film, TV and consumer goods, this has led to some tremendous hits: "Gone With the Wind," "All in the Family" and Jello Pudding. 

But of course, even bus-tour slots players hit jackpots. Over the long haul, nobody gets close to even 51%. The misses vastly outnumber the hits -- and the misses aren’t near misses. They are bellyflops: Ishtar, "My Mother the Car" and Touch of Yogurt Shampoo.

Although, maybe, never again. In online TV programming, at least, pinkbelly may be a thing of the past.

The world of streaming video doesn’t depend on cigar-chomping tyrants, taste-sensing dowsers or even the volumes of Simmons demographic indexes that for so long were the tools plumbing the consumer mind. Because now even the most unibrowed balagula can figure out what the public -- or at least, some public -- actually wants.

“Media has been held hostage by the notion that the only way to understand the audience is through its demographics,” says Johanna Blakley, managing director and director of research at USC’s Annenberg Norman Lear Center. This was the currency of the industry and still is, but it was really just a shortcut.

“There was this sense that you couldn’t possibly understand what individual members of mass audiences were interested in, but, if you knew a little bit about their age, their gender, if you knew how much money they made, their race, you could make some assumptions about what their interests might be, and this interests might well be associated with their purchasing behavior. That is just not the most elegant solution to finding an audience.“

No, it’s like butchering a steer with a sledgehammer, and yet hundreds of billions of dollars changed hands based on data little more telling than a ZIP code.

“I think the reverse is going to start happening, where suddenly we’re going to have the computing resources in order to figure out what audiences exist out there, groups of people who have very similar patterns of interest… and let me tell you, they don’t live in the same ZIP code. It’s going to be a kid in Bangladesh who has the same interests as a kid in Queens who has the same interests as an old guy in North Dakota who went up there to start fracking.”

She doesn’t really have to prognosticate on this. Blakley’s actually retro-nosticating. She has seen the future, and it is Netflix for the past two years. Turns out that Reed Hastings' magical algorithm is handy not just for predicting what movies in its acquisitions library you are most likely to enjoy, it can deduce what new production might be just the ticket, too.

“'House of Cards' was a safe bet for them, to spend $100 million to get that show off the ground because they already knew that there was a Venn diagram that already overlapped fans of David Fincher movies, fans of Kevin Spacey and people who had seen the original BBC "House of Cards." So they knew there was an audience there. So they knew it was less of a gamble for them.”

"Same with 'Orange is the New Black'," which producer Jenji Kohan shopped around to premium cable with no success before heading to Beverly Hills to pitch Ted Sarandos at Netflix.

“I know she pitched it to HBO and Showtime and they said absolutely not. She went to Netflix and they decided to see if there was any audience for it and they said, “Oh my God…there is!” Make it!”

Why? Because the data indicated a pent-up demand for all these qualities: Provocative female lead, female ensemble cast, prison drama, dark comedy and -- woo-hoo  -- lesbian sex.

A few miles west in Santa Monica, Amazon Studios also has the benefit of collaborative filtering. That’s how Jeff Bezos knows that people who like lox also like bagels, people who buy propane stoves also buy sleeping bags. But Amazon not only crunches the numbers to deduce demand, it crowdsources both content and feedback at every stage of development -- sort of an ongoing, massively-multiplayer focus group.

“The goal at Amazon Studios is to be the most audience-focused television network there is out there,” says Joe Lewis, head of original programming at Amazon. “It’s a network that is solely based on making content that customers love.”

So, for instance, if you download a pilot to watch, Amazon watches you watch…keeping track of pauses, rewinds, abandons and so on. And, of course, it has the same ability for actual offerings -- such as "Alpha House" -- you can stream from Amazon prime today.

As I sat with Lewis, listening to the details of these capabilities, I reminded him of the famous William Goldman wisecrack, the dismissive quip about the so-called tastemakers. “Nobody knows anything,” Goldman declared. Of course, that was then. This is now, and Joe Lewis smiled.

“Nobody knows anything,” he said. “but if you listen you might hear something.”

Recommend (5)
2 comments about "Beyond Your Wildest Streams, Part II: Data Is The New Czar".
  1. Mark Klein from Loyalty Builders LLC , March 10, 2014 at 11:23 a.m.
    Garfield is right, data is the new czar, but he misses the really big opportunity. Analyzing transaction data at the individual customer level (what people have already purchased) will let you predict what customers will buy next, even stuff they have not yet purchased (cross-sell). This means that when your company sends out a million emails, each one can be different, offering different products to different customers. Companies doing this today are seeing spectacular returns.
  2. Theresa M. Moore from Antellus , March 10, 2014 at 4:32 p.m.
    Of course, why Amazon should try to monitor watchers leaves a creepy sensation down my spine. Amazon would love to anticipate and provide value to its customers, but fails in every regard. Tailoring films and other content according to the lowest common denominator will be its downfall, which would not be a bad thing in everyone's estimation. Pushing products on people means that Amazon is just like any other bad executive -- ignorant of the reality on the ground, and unable to see what is right in front of it. One of the many reasons I don't work with Amazon.