Why Netflix and Amazon Algorithms Are Destroying the Movies
A long time ago, I thought that the worst interface I had ever seen — after most film festival websites, that is — was to be found on Time Warner (now Spectrum) cable. But today, what could be worse than Netflix and Amazon for finding movies?
Think of any new documentary or arthouse film. Don’t search for the title; look for it by browsing — you know, like 99% of customers do when they turn on their internet connected TV. Odds are good that you’ll not find that film until you’ve swiped, or toggled or (depending on your device’s interface) clicked through 10-20 screens.
Last night, I decided to look for the Grateful Dead documentary “Long Strange Trip,” which Amazon released last month. It was an acquisition; the company wanted to make money from it. It was a long, strange trip indeed (sorry), as it took me 20 screens or so of swiping before I came upon the movie. I came across numerous older documentaries, a shit-ton of crappy docs about blenders, ex-porn stars and other junk, and then the Dead.
If I wasn’t dead set on finding it, I would have given up and ordered something else. And that’s a problem for any video/film content that isn’t a blockbuster or a TV show. This doesn’t bode well.
Whether it’s Netflix or Amazon, the problem is the same: It’s the algorithm. Silicon Valley will tell you the algorithm is god. Bullshit. Amazon knows my entire purchase history — not just films, but books and, well, everything else. But none of the titles their algorithm was pushing at me had any relation to anything I’ve ever watched or wanted to watch. (There was a documentary about a stand-up comedian; I’ve watched one of those.)
But the algorithm is supposed to be smart. It’s supposed to know what I want before I want it, and it never does. Same with Netflix. Judging by their algorithmic offerings, I only like TV shows, stand-up comedians, and I don’t want to go anywhere near subtitles, documentaries, or american indies.
But this isn’t true. Not only is that what I like to watch — I’ve watched them more than TV shows on Netflix, unless the platform counts every episode of “House of Cards” as separate movies (actually, I bet it does). iTunes isn’t much better, although if you type a genre in there (instead of browsing), you do get some diversity. More people go to iTunes with a film in mind, while they browse more on Netflix and Amazon Prime. Vimeo’s discovery interface is horrendous, too, but I won’t pick on them because they’re way smaller and offer up more indie content by the nature of their business model. I’m not mentioning other VOD services because, well, no one uses them.
Shut up, old man, you might say. But this is a huge problem for quality films whether they’re docs, indie, foreign, or classics. If the algorithms can’t serve these up to me, I guarantee they aren’t serving them up to anyone else. Netflix, in particular, seems to be pushing them further down the list — and, in the process, making it hard to find movies at all. As we all know, arthouse films, especially docs, are bombing in theaters if they make it there at all. Many go straight to digital, maybe with some touring. But most film watching these days is online, and if you can’t be found there, well, you don’t exist.
Sure, you can do your own marketing, send your fans there, and they can find you via direct link. But most humans sit down, scan and buy. And they tend to watch what’s on the home screen or within a few clicks. I know this because I’ve spoken with staff at some of these places off the record, and they admit that their data shows that people don’t search too far (they also barely use their queue, and tend to put aspirational films there that they hope to watch but never do).
Setting aside the fact that Netflix and Amazon are buying fewer films overall (that’s a problem too), and fewer documentary and arthouse films every year, as told by every sales agent, these guys aren’t even trying to push the ones they do buy. And when the algorithm buries those films,it becomes a vicious cycle, with Netflix deciding they don’t work and let’s buy even fewer of them. As filmmakers and film lovers, we end up just where we were with cable pre-Netflix — the illusion of choice that is so vast we don’t realize how much we’re actually missing. So, where I used to always worry that this whole net thing would end up with just giving me a better TV, now I’m worried I won’t even get that. Forget jet-packs…
Last week, I posted about the need for more funding around discovery. Kent Bye, a very smart man who mainly works in VR now, posted this comment, which I quote here in full. It’s about YouTube, but it’s interesting for this too:
YouTube has 300 hours of video uploaded every minute, which amounts to about 49.28 years of new content per day. Here’s a really insightful video from a popular YouTube creator breaking down the recent changes of the YouTube algorithm to preference new daily videos over older and more higher quality ones. Content creators who depend upon that algorithm for revenue are at the whims of YouTube’s changes, and those who create content optimized for the algorithm produce shoddy, low-quality, clickbait that has a monetary incentive to figure out new ways to game the system, produce glossy and deceptive thumbnails, and keep people hooked in a loop of the latest drama. The current system is producing a lot of empty calories.
As Ken points out, we are fucked, we are fucked, we are fucked (WAF). WAF1 is superabundance. Your film is not just up against other films. It’s up against every video being uploaded. Heck, if we stick just to film, there are estimates of 50,000 unique titles (or more) being submitted to festival submission sites annually.
WAF2 is the algorithms are not set up to help me, or you, find the content we want; it’s designed to glue us to each service. “House of Cards” is great TV, and Netflix knows snacking on that addictive in-house production will bring you back more often than “The Turin Horse.”
That’s WAF3: Expect more episodic TV and less worrying about quality films (much less esoteric films). Another time, we can argue about how episodic series have supplanted film in the public conversation and whether that’s a good thing. (I watch a lot of it too.) But WA4 is we can’t sustain independent or fringe voices if the only two buyers who matter anymore aren’t buying our films. The problem will move from one of discovery to one of absence.
That’s why I think we need to focus more money on discovery now, and less on creation. This is some dire shit, and my hunch is that solutions have to come from outside the system. The disruption won’t come from a competitor to Netflix (you’d need a billion dollars just to begin to compete). Just like the indie doc world has built a support network for funding (Britdoc, Sundance Doc Fund, Just Films, pitch markets, forums, etc.), we need to come up with solutions to help curate better, help people discover and remember films (since they often hit Netflix 90-180 days after theatrical), and help them to find a diversity of films.
I hope to have more thoughts on that soon, and hope to hear them from you.
Brian Newman is a film producer and consultant who runs Sub-Genre Media.
