Wednesday, April 26, 2023

Minor Post #5 by Mike Goemaat

I'm excited to talk about Global TV and Post-TV together for our final class because I believe there are important connections to tease out, specifically as both topics relate to streaming. Continuing with my semester-long fascination with Netflix, I was impressed how, writing in 2004, Lisa Parks already anticipated the way that television would become increasingly personalized as we moved away from watching whatever was currently on air. She writes: "Moreover, by removing the spontaneity of the practice of channel surfing, this model of flexible microcasting allows programmers to determine more accurately where and when viewers are in the media landscape" (Parks 136). But even she could not have foreseen how prescient her line about "programmers determining" our choices would become. 

Luckily, scholar Amanda Lotz has written extensively about this very subject. In her article "In Between the Global and the Local," Lotz discusses the ways that Netflix asserts is multinational status and is able to program for niche audiences while providing a global service. Read as a present-day version of microcasting, Netflix utilizes its tremendous amount of user data to offer personalized "recommendation clusters." And these clusters are wildly successful at predicting what users will enjoy watching, so much so that Netflix asserts that "80% of Netflix viewer activity is a direct result of personal recommendations" (Behind The Scenes of The Netflix Recommendation Algorithm). This is, quite frankly, an absurdly high hit rate for a content provider to have. But as Lotz points out, "Netflix targets subscribers based on tastes and sensibilities that are often not sufficiently popular to be address by services aiming for a national 'mass' audience [...] but SVOD viewers derive value at the individual, or rather household or subscriber level" (Lotz 207). Although Netflix is servicing 230 million users in 190 countries, and is cognizant of creating "hits" on its platform, its direct-pay model means that it just needs to offer a piece of content that will keep the individual subscriber happy. This allows it to take risks, however esoteric, to appeal to niche audiences whose tastes belong in specific "recommendation clusters" and even approach its subscribers as their own "transnational clusters of tastes and sensibilities" (Lotz 207). It is a fascinating example of the way that, even on the most popular and widely used streaming platforms, we are more interested in personalizing our TV experience than ever before. 

(Final note: I find it extremely interesting that Netflix's website has a "help" page that explains how their recommendation system works, including tell you all the different data points that are fed into the algorithms to offer more and more accurate suggestions. It seems the early seasons of Westworld were right along: the devil is in the data.)

Check it out here: https://help.netflix.com/en/node/100639 




 


1 comment:

  1. Comment/Minor Post #4 from me: I really appreciated your final note here, Mike, regarding Netflix's recommendation system. There's something with Netflix and other major data aggregators and tech companies that produces a tension between transparency and opacity -- they'll provide you a breakdown of how "the algorithm" works, but will also guard said data very carefully, without enabling a clear glance into that sphere.

    ReplyDelete