> If the TV networks had a clear picture of viewership, they would be more effective at negotiating licensing deals.<p>The TV networks are still stuck in the mental space of thinking of viewership count as a relevant metric because for them more viewers equals more value for ads.<p>But for Netflix the revenue they make is not advertising dependent so it doesn't matter whether a show has high viewership as long as it helps Netflix retain subscribers. A small show with a cult following that helps Netflix retain a subscriber for a longer period of time can be worth more money than a big box office hit that brings in a burst of short term viewership.
"In the last 30 days."<p>I mean this kind of makes sense, doesn't it? Netflix releases episodes all at once. Many people are going to binge or watch them very close to that release date.<p>Their survey was conducted in a lull for Netflix OG content.<p>Also when you can watch anything on demand, does counting what had the most eyeballs in the last N days mean much? Isn't "how much a show is watched at all" the better question?
If you want to advertise on Netflix you need product placement in the shows. (I think there would be a revolt if the splice in old-school advertisements.) Having some numbers on eyeballs helps justify the ROI for sponsoring the creative development of a new show.
Are all Netflix Originals produced like House of Cards? If all the Originals were produced aided by machine learning methods, can we say that machine learning methods are not that effective?