At what point does ad personalization become profiling?
Weve known for ages that Netflix uses personalized imagery to sell its wide array of programming in different ways to different users. When using Netflix, for instance, some people see a nice still of Claire Foy as the platforms main image for The Crown; others see a cheerful bunch of corgis. As detailed last year in a blog post, Netflix selects these images for each user based on their viewing data, selling its projects differently depending on whos looking. But recently on social media, a few black users have reported being shown misleading imagery on Netflix—images that feature black actors who only briefly appear in the film or show theyre being used to advertise.
Last week, writer and podcaster Stacia L. Brown posted one example on Twitter: a promotional image for Like Father, a Kristen Bell and Kelsey Grammer vehicle, featuring two black actors who only briefly appear in the film. Brown also experienced the same thing with other films on Netflix, including Set It Up and Love Actually. Although users are torn about how insidious this practice is, one consensus seems clear: the bait and switch can get frustrating.
In a statement, Netflix essentially denied the charges, emphasizing that it does not collect or use demographic data to determine what art to promote: “We dont ask members for their race, gender, or ethnicity so we cannot use this information to personalize their individual Netflix experience,” a representative told Deadline. “The only information we use is a members viewing history.” The streaming service has previously explained its strategy by saying that it sorts users into “taste communities,” determined by tracking the types of content a given viewer watches.
If that statement sounds vaguely familiar, you might be catching echoes of a similar argument Facebook used a couple of years ago, while defending a marketing tool meant to help advertisers connect with or exclude members of certain communities. Facebook said it did not overtly profile its users based on their demographic or census data, but instead categorized them based on their apparent “ethnic affinity”—such as their declared interests and languages spoken. Last year, after much criticism, the features ability to exclude certain groups was disabled.
Get Vanity Fairs HWD NewsletterSign up for essential industry and award news from Hollywood.Laura BradleyLaura Bradley is a Hollywood writer for VanityFair.com.
[contf] [contfnew] 
Vanity Fair
[contfnewc] [contfnewc]
