As the Internet scratches its Hydra-head over Netflix’s announcement that it’s splitting off its DVD-by-mail rental service under the unlovely heading of “Qwikster,” Tim Lee tweets that Bill Gurley’s speculation is the most plausible explanation he’s seen for a move consumers seem to be universally panning:
So here is what I think happened with Netflix’s recent price change (for the record, I have no inside data here, this is just an educated guess). Netflix has for the past several years been negotiating with Hollywood for the digital rights to stream movies and TV series as a single price subscription to users. Their first few deals were simply $X million dollars for one year of rights to stream this particular library of films. As the years passed, the deals became more elaborate, and the studios began to ask for a % of the revenues. This likely started with a “percentage-rake” type discussion, but then evolved into a simple $/user discussion (just like the cable business). Hollywood wanted a price/month/user.
This is the point where Netflix tried to argue that you should only count users that actually connect digitally and actually watch a film. While they originally offered digital streaming bundled with DVD rental, many of the rural customers likely never actually “connect” to the digital product. This argument may have worked for a while, but eventually Hollywood said, “No way. Here is how it is going to work. You will pay us a $/user/month for anyone that has the ‘right’ to connect to our content – regardless of whether they view it or not.” This was the term that changed Netflix pricing.
What still has Tim puzzled is: Why would Hollywood try to insist on that?
It does sound bizarre at first blush, but I think it makes a certain amount of sense when you think about substitution effects, and the real reason people choose to own movies rather than rent them.
The simplest answer might be that it’s a straightforward function of how many times you expect to watch a movie. If the price of purchase is lower than the number of expected viewings multiplied by the rental cost, it will be cheaper to buy.
Except I don’t think that adequately captures why people buy movies. I have a reasonably large DVD and BluRay collection—somewhere in the ballpark of 100 movies and TV seasons. There’s a handful I never seem to get sick of—The Big Lebowski, The Dark Knight, Brazil, episodes of Firefly—and maybe 20 all told that I’ve watched (or am likely to watch) five or more times, which I’d guess is the average rent/purchase break-even point. If that were the only consideration, the other 80 might seem like irrational purchases. Quality is another factor—a BluRay still looks and sounds noticeably better than a streamed movie on a big screen with good speakers. But that’s not all there is to it either.
Perhaps the biggest advantage of purchase over rental is having ready access to a collection. That is, if I’m lounging around with my girlfriend on a rainy Sunday evening, and decide we want to watch a movie right then, we have a pool of 100 movies to choose from without either of us having to truck out to a store or rental kiosk. Maybe we’ll never get around to watching Apocalypse Now or Repulsion five times, but the point is that we could watch them anytime.
Netflix streaming changes the calculus by giving you a huge pool of instantly-available movies without requiring you to own a large collection. For the 20 or 30 favorite movies, I might want to own them anyway, especially if they’re not necessarily going to be perpetually available to stream. But the “long tail” of my collection consists of a lot of movies that I own because I want to have the option of watching something pretty good in that genre anytime, not necessarily that particular movie. With Netflix available, I’m more likely to buy the 30 percent of favorites, and then assume they’ll have something acceptable streaming when we want to watch something new. There’s no need for a big collection of physical discs.
Now it’s easy to imagine the studios initially seeing streaming as primarily displacing rentals. That’s pure gravy for them, because the First Sale Doctrine means they’re not making any revenue from rentals after the initial purchase. Then—whoops!—they realize it’s actually displacing sales, because ready access to a pool of movies is actually a pretty good substitute for ownership. So it’s not so mysterious that they might suddenly want a fee based on the total number of subscribers with access to a streaming film. Because some percentage of those, even if they don’t end up watching a particular movie more than once or twice, will think: “I could buy it, but why bother if I can just stream it whenever I feel like seeing it? Even if that one movie gets pulled from streaming, there will be plenty of others about as good to choose from.” The mistake—perhaps natural for folks who spend their time making and marketing individual films—was not seeing that consumers often aren’t so much interested in watching some particular movie as they are in the ability to watch something. Just as many people spend hours “watching TV” rather than watching any particular show, people often just want to “watch a movie”—de dicto, rather than de re, as the philosophers say—or rather have the option to watch any one of a number of movies, more than they want to see any particular one.