The streaming service’s clickbait problem threatens to ruin it.
I’m on my couch, scrolling, scrolling, scrolling. As I scan my Netflix page, which rectangular tiles jump to the foreground of my attention? Well there is What is below, with his promotional photo of an incredibly handsome man emerging, dripping, from the surface of a lake. There are Sexify, with her promotional photo of a woman’s face seemingly captured at the height of passion. Or maybe I should watch Lucifer, his star staring at my shirtless soul, his chest so strangely hairless that it resembles that of a video game character.
The answer to What is below maybe an aquatic geneticist with a Superman jawline, but other mysteries aren’t answered by Netflix’s most popular titles. For a while this spring Why did you kill me? scrambled for space on the Top 10 list with Who killed Sara? Why made you kill me? who made kill Sara? Is it the girl in the yellow bikini? Where is she Sara? There is only one way to find out: Click on.
In these late pandemic days, there is a dizzying familiarity with the screen where we make most of our entertainment decisions. To anyone who’s ever written titles for online media, Netflix looks familiar. As competitors begin to pinching the heels of the streaming giant, Netflix seems to be flirting with the tantalizing tools that web publishers discovered a decade or more ago: the curiosity gap, the sexy thumbnail, the deceptive image. A homepage is a homepage, after all, and these days Netflix has discovered clickbait.
Netflix has always been committed to getting people to click on a show, and fast. The network’s own research shows that users consider each title for 1.8 seconds, and if users can’t find anything in a minute and a half, they’re gone. In the battle against “paralysis of analysis“, the network is constantly experimenting with new ways to make homepage elements as attractive as possible.
And Netflix recognizes that works of art – the thumbnails they feature – are crucial to that decision-making. Since 2017, Netflix has personalized his tiles depending on a user’s viewing history, a change from their original quest to find the unique sticker to attract them all. This means (to take the example of Netflix) that if you watch a lot of romance, the company may offer you a thumbnail for Goodwill hunting which features Matt Damon and Minnie Driver kissing. If you like comedy, your Goodwill hunting the sticker may feature Robin Williams. (More doubtful, as outlets such as Wired have reported, if you are black you will end up seeing black artists in your thumbnails, even if they are relatively minor characters in the movie or series. I see Liam Neeson in my Love, in fact thumbnail; others might see Chiwetel Ejiofor.)
The service’s homepage and the shows acquired by the network have long been ruthlessly optimized to appeal to very specific audiences. When new content appears on your Netflix homepage, it’s often cheekily similar but not quite related to the content you’ve enjoyed before. After the success of the generically titled comic Money theft, Netflix recently launched the new, even Following generically titled hit Robbery—A sticker depicting, to make sure nothing is overlooked, a woman wearing only her underwear and a cascade of $ 20 bills.
Sometimes the serve stretches so far that it seems to be on the verge of total orientation error: take Ragnarok, a Netflix title illustrated with an image of Thor’s mighty hammer. Click through to this new Marvel Comics Universe series and you’ll end up watching a perfectly enjoyable Norwegian show on Norse mythology. And the company’s reliance on custom artwork means it often seems to be playing 4-dimensional chess with recommendations. For some reason (maybe the action movies I recently streamed?) My homepage shows The great Lebowski with a photo of an angry John Goodman pointing a gun, as if the Coen Brothers’ comedy was actually some sort of revenge thriller.
I mean, whatever. As long as there have been movies, there have been marketers plotting to get people to watch them. But this spring and summer things have looked a little different at Netflix. The company seems more cheeky in its strategies, more willing to promise something and then absolutely fail to deliver, often using familiar headline tricks from the social web. That’s what the click bait is: get someone to click, then offer them something other than what the title made them want. You ask a question, but you do not answer it. You promise satisfaction, but you leave the user unsatisfied.
You can lazily click on a thumbnail to Gender / Life– with his gorgeous, sweaty panting central figure – not because you expect quality, exactly, but because you expect some type of aerated waste: Skinemax for the 21st century, with just a little bit of character development to help it flow smoothly. It might not even be a big deal if it doesn’t Well. It just needs to satisfy the urge you felt when you clicked on it.
Clickbait works, at least when users first encounter this.
But many of Netflix’s recent clickbait hits don’t keep their promises, implied or explicit. Gender / Life is “more soapy than sordid,” as Slate’s Karen Han put it, streaking her episodes of vanilla sex scenes but making this main character’s moral dilemmas so boring the scenes are impossible to enjoy. What is below offers neither great sea monster sex nor thrills. Cripes, you can keep clicking for a whole season of Who killed Sara?, 400 minutes of melodrama, only for … never find out who killed Sara! Tune in for season 2 I guess.
Living the last decade in the news online has made me a little wary of this particular kind of all-in-one, cattle-free marketing. Take these headlines-as-questions, for example. In their frankness, they ask for an answer and bring back the headlines of the ‘curiosity gap’ that plagued the media in the mid-2010s, as every website attempted to replicate Upworthy’s success: Someone killed Sara. You will never believe who it was. Likewise, the bland and sexy miniature images are reminiscent of the most basic of chumbox, those sometimes disturbing programmatic ads that plague the bottom of websites (including, sometimes, Slate) that need to make a little more income from their real estate online.
It seems unlikely that Netflix’s clickbait trend in 2021 reflects a decision by the executive; it’s probably a confluence of a few surprise hit titles and a machine learning algorithm that just uncovers that particular flavor of human gullibility. Because clickbait works, at least when users first encounter it. It worked when Upworthy and other sites started playing with it in the early 2010s, and these titles were all over the network. “one hundred percent goalPopularity lists suggest they are working now. And once an algorithm sees something working, it goes crazy to replicate those results as much as possible.
The question is: what will Netflix do about its algorithms that uncover the lowest instincts of humans? It is proven that even now their internal metrics change to emphasize not if you click on a show, but if you stick to it. Since 2019, Netflix’s definition of whether a show has been “watched” was just that a viewer had seen two minutes of it. This is an obviously bad metric; two minutes is not even enough to understand that Chris Hemsworth is not there Ragnarok! But I’ve spoken to several people familiar with the metrics that Netflix uses internally and shares with creators, and they’ve suggested that the grip – that viewers go 80% of the way through a series, for example – becomes of more and more important than if the viewers click on it. the first place. “Things on Netflix are always changing,” one show creator told me, and “we don’t get a lot of information,” but that creator has been made clear that their show has been renewed not at because of a large number of people clicking, but because a high percentage of people watched all season.
And it’s good ! This is the kind of policy that will hopefully nip Netflix’s bloom of clickbait in the bud. Believe me, on behalf of online media: it’s great at first to get people to click on girls in bikinis by asking provocative questions, but you’ll never guess what will happen next.