Will Netflix's Use of Data-Mining Change the TV Production and Distribution Paradigm?

A November 29th, 2012 article in Wired by Roberto Baldwin reveals the enormous gamble that Netflix is taking, betting the future of the company on the ability to predict audience response for new big-budget programming.

For example, Netflix has reportedly bid over US$100 million and committed to two full seasons (26 episodes) of Media Rights Capital’s new drama series House of Cards, executive produced and directed by David Fincher and exec produced by and starring Kevin Spacey. According to Roberto Baldwin: "The company knows how many people are watching Kevin Spacey and David Fincher movies and it knows how many viewers watch political thrillers. If that audience is large enough, getting exclusive access to House of Cards makes sense."

The key question (that has many Hollywood veteran's smirking in anticipation) is whether the Netflix algorithm is superior to the old school decisions made via the guts of programming executives: "Even a TV show with an astounding cast and big-name writer or director can fail."

But Netflix knows that the market for existing content is tightening up - and Netflix has a proven history of matching online viewers with content using sophisticated algorithms.

Is the current Netflix Cinematch algorithm the secret to becoming a competitor to other subscription services like HBO?

So far HBO is not allowing online-only subscriptions. So Netflix sees an opportunity to provide premium content via an $8 per month streaming deal - without head-to-head competition with HBO or other premium original content providers.

But key questions remain unanswered:

Can Netflix accurately predict audience response to NEW content based on the prior viewing habits of their 27 million subscribers?

Can Netflix translate knowledge about audience preferences into a competitive advantage over all the other sources of motion picture entertainment (including YouTube, Hulu, Google Play, Amazon Prime, HBO, Showtime, and all the other existing TV outlets)?

What happens if HBO or Showtime start offering online-only subscriptions? Will users still pay for Netflix if other premium outlets start offering similar fare?

What will happen if Amazon Prime and Google - with their much larger reserves of capital - start producing original content and bidding against Netflix?

And, as noted in a Nov 30th, 2012 post to indieWire, the ultimate test of Netflix's algorithmic model will not be whether series with big stars succeed: "Where Netflix's use of viewer data will get really interesting is in the next round or two of programming (assuming they happen), when projects could be more about serving specific niches than latching on to big names."

UPDATE February 3rd, 2013: According to a February 3rd, 2013 post to The Atlantic Wire, "[t]he real problem for Netflix is that their subscription revenue is not growing as fast as their content costs. Michael Pachter, an analyst with Wedbush Securities, told Bloomberg News' Cliff Edwards. "Netflix will continue to generate negative cash flow going forward, driven by the company’s ever-increasing streaming commitments.""