I checked out the Salon article
that quote was taken from and it's even funnier and scarier than I thought.
Correlating the raw numbers of Kevin Spacey fans who also like David Fincher and have a fondness for British political dramas is just the beginning. Netflix knows enough about what you are watching to judge specific aspects of content as well. Last summer senior data scientist Mohammad Sabah reported at a conference that Netflix was capturing specific screen shots to analyze in-the-moment viewing habits, and the company was “looking to take into account other characteristics.”
What could those characteristics be? GigaOm’s report of the Sabah presentation speculated that “it could make a lot of sense to consider things such as volume, colors and scenery that might give valuable signals about what viewers like.”
Wow, that's getting down to minutae pretty quickly, but what about the bigger picture. They're making a basic mistake if they're assuming that the whole universe of "what people want" can be based on what already exists. How about what doesn't already exist, but should? Market research can't substitute for creativity.
And even if they're just looking at what already exists in their library, the number of possible connections is staggering. Why spend $100M on the House of Cards-Spacey-Fincher connection and ignore all the other dots that could be connected? I'll bet the Star Trek-Ben Browder-Josh Whedon axis is pretty strong if you thought to look for it.
And how do they know any given connection is meaningful? Maybe a gangster drama starring Dustin Hoffman and directed by Stephen Soderberg would have done just as well, even though there may be no particularly strong relationship that can be found in the data, but those three elements would click for ineffible creative reasons.