Since 1998
-----------------

 the beat of digital culture
home | archives | about us | feedback

weblog:
Daily Relay

Tracking digital culture and emerging technologies

special section:
Mindjack Film
Fresh thinking on current and classic cinema

shop:
Mindjack T-Shirts
Only $15US
shipping included

Google
Web
Mindjack

Mindjack Release
Sign up to receive details of new issues

 

May 13 , 2005 | PART ONE: HYPERDISTRIBUTION

October 18th, 2004 is the day TV died. That evening, British satellite broadcaster SkyOne — part of NEWS Corp's BSkyB satellite broadcasting service — ran the premiere episode of the re-visioned 70s camp classic Battlestar Galactica. (That episode, "33," is one of the best hours of drama ever written for television.) The production costs for Battlestar Galactica were underwritten by two broadcast partners: SkyOne in the UK, and the SciFi Channel in the USA. SciFi Channel programers had decided to wait until January 2005 (a slow month for American television) to begin airing the series, so three months would elapse between the airing of "33" in the UK, and its airing in the US. Or so it was thought.

The average viewer of the SciFi network is young and decidedly geeky. They are masters of media; they can find ways to get things they shouldn't have. Thus, a few hours after airing on SkyOne, "33" was available for Internet download. No news there.

A new peer-to-peer file sharing technology, BitTorrent, was employed to share the quarter-gigabyte audiovisual files of "33". Unlike older forms of internet downloading, where too many requests for the same data can clog up internet links and send servers crashing. BitTorrent distributes files more and more efficiently, as more people join the hunt for the data. Everyone looking for bits of a file - say, an episode of Battlestar Galactica - shares the pieces they've already located with anyone else who doesn't already have that piece. Since the pieces are scattered randomly among all the users who want the data, there's a lot of to-and-fro between the users; rather than being a request for one copy of one file on one server, it's as though many hundreds of hands are copying and exchanging playing cards. You may start out holding only the Ace of Hearts, but soon enough you'll have a full deck.

This is a form of peer-to-peer file sharing known as "swarming": all of the peers in a swarm share the portions of the data they've already received. And, as the Chinese proverb goes, "Many hands make light work." BitTorrent transforms the creaky and unreliable technology of audiovisual distribution, making it fast and hyper-efficient. BitTorrent creates the conditions for something I've termed "hyperdistribution" - a distribution channel which is even more efficient than broadcasting.

That has certainly been the case with Battlestar Galactica. The British aficionados of the series provided torrents for each episode within a few hours of each broadcast. Many fans in the US picked them up and watched them; so did many people in Australia.

While you might assume the SciFi Channel saw a significant drop-off in viewership as a result of this piracy, it appears to have had the reverse effect: the series is so good that the few tens of thousands of people who watched downloaded versions told their friends to tune in on January 14th, and see for themselves. From its premiere, Battlestar Galactica has been the most popular program ever to air on the SciFi Channel, and its audiences have only grown throughout the first series. Piracy made it possible for "word-of-mouth" to spread about Battlestar Galactica.

Just two months ago, we saw something very similar happen, again with a beloved series, the BBC's Doctor Who. After a hiatus of almost two decades, the BBC cast Christopher Eccelston in the role of the Doctor, and set the show to premiere on the 24th of March. A few weeks before the air date, an "unfinished" version of the first episode of the new series leaked onto the internet through the BBC's production partner CBC. Hundreds of thousands of Doctor Who fans downloaded the episode, wanting a preview of this new version of the nearly-immortal Doctor. The BBC were publicly outraged, but there's a strong sense that this act of piracy, while not officially sanctioned, was unofficially encouraged by BBC. It certainly created a groundswell of interest in the series, allowing people to "try before they buy," and probably increased program viewership. (The episode drew 10.81 million viewers to BBC1, which is among the highest ratings Doctor Who has ever seen.)

Audiences are technically savvy these days; they can and will find a way to get any television programming they desire. They don't want to pay for it, they don't want it artificially crippled with any digital rights management technologies - they just want to watch it. Now. This is the way that half a century of television and a decade of the Web has conditioned them to behave. We can't really complain that audiences are simply doing as they've been told. It is pointless to try to get them to change their behavior, because, in essence, you're fighting against the nature of television programming itself, the behavioral narrative which grew out of our relationship to the technology. We all understand that this piracy is technically illegal, technically a violation of copyright; but we're in a hell of a bind if we're telling the audience to "sit down, shut up and do as you're told" when it comes to television viewing. The audience won't do as they're told: they'll do as they've been taught, and that is another story entirely.

Still, piracy presents us with an economic problem: how do producers get paid for the programs they create when audiences disintermediate the distribution channels through which producers get paid for their programming? The economics of television production, as practiced for the last fifty years, are very straightforward: producer (or perhaps the producer's distributor) sells the program to a broadcaster. Broadcaster sells commercials to advertisers. Everyone gets what they want: the producer gets enough money to cover his costs, the broadcaster gets money to cover his costs, the advertiser gets some attention from the audience, and the audience gets the program.

Widespread piracy of television programming has short-circuited this process, connecting the producer directly to the audience. As yet there are no viable economic models connecting the television producer directly to the audience. Industry pundits talk about audiovisual downloads through some system like Apple's iTunes Music Store, and perhaps we'll see something like this in the near future, but this works against the simple fact that people do not expect to pay for television programs. People will pay for movies, when they choose to pay for movies, but they won't pay for television programming. Not if they can get it for free. The audience is not at all involved in the economic value chain of television production; that's been the rule for a half-century. It's reasonable to presume that any attempt to change the economic behavior of the audience is doomed to failure.

Cable and satellite broadcasting presents something of an argument against this assertion insofar as people do pay for these services. But in these cases the audience is really purchasing choice. (Los Angeles has at least 20 broadcasters, and, despite this, has a thriving cable and satellite broadcast market, because people want even more choice, and are willing to pay for it.) Hyperdistribution has extended this choice to anyone with an broadband connection — extended it well beyond any possible offering by any cable or satellite broadcaster. Can these industries possibly compete against the nearly infinite range of content offered on a broadband connection?

Now we have a paradox: the invention of an incredibly powerful mechanism for the global distribution of television programming brings with it a fundamental challenge to the business model which pays for the creation of the programs themselves. This is not at all BitTorrent's fault: the technology could have come along a decade ago, and if it had, we'd have stumbled across this paradox in the 1990s. This is a failure of the value chain to adapt to a changing technological landscape — a technological desynchronization between producer and audience. Once again, there's no need to find fault: things have changed so much, and so quickly, I doubt that anyone could have kept up. But the future is now here, and everyone in the creative value chain from producer to audience must adapt to it.

This presentation outlines one economic model — actually more like a family of models — which connects television producers to their audiences through an hyperdistribution strategy, one which doesn't require any change in the audience's economic behavior. This, I believe, is the surest path to success for any new economic model; without audience acceptance, any model will inevitably fail, and while this model is not guaranteed to be successful, it seems to face fewer roadblocks to acceptance than other models which have been proposed.

Television broadcasters owe their existence to the absence of substantially effective competition. When you're dealing with real-world materials that are in naturally short supply - whether diamonds, oil, or broadcasting spectrum — a cartel can maintain and enforce its oligopoly. But when you're working with media, which exist today as digital ephemera, bits that can be copied and reproduced endlessly at nearly zero cost, broadcast oligopolies are susceptible to a form of "digital arbitrage," which can hollow-out their empires in an afternoon. Hyperdistribution techniques are more efficient than broadcast networks for television program distribution.

Now, before you presume that this is all so much future talk, that maybe, someday, people will be downloading television programs from the Internet, know this: that someday has already come and gone. Per capita, Australians are the most profligate downloaders of television programming in the entire world, followed closely by the British. While the Americans lag behind, they're still on the chart, in third place. The sea change has already taken place - undoubtedly sped along by the monopoly position of the commercial broadcasters, who, in many cases, act as barriers rather than conduits for television programs. If a commercial broadcaster doesn't show a program, or delays it for years, that's no longer of concern to television audiences: they'll just download it from the Internet.

This trend is only going to accelerate with the uptake of broadband throughout the world, progressively hollowing-out the commercial broadcasters until they have returned to their roots: television as a live medium. The only types of programming unsuitable for hyperdistribution are those which are broadcast live: news, event and interactive programming, and sport. Since these are all widely popular, it's not as though the commercial broadcasters will collapse. But their business models will change, because their cash cows are fleeing the paddock.

The pervasive culture of TV downloading leaves the producers of pre-produced television programs high and dry, receiving nothing of value for their work. But is this really true? The absolute, basic motivation of a TV producer is not money — though money is needed for production — but to gain and hold an audience's attention. TV producers want their programming to be watched as widely as possible — by everyone. That's what they care about, and that's all they care about, because, with viewers, everything else takes care of itself: audiences equal money.

This assertion seems so basic, so fundamentally essential to the economics of television, that it's very hard to understand why anyone (other than a broadcaster being cut out of the value chain) would get upset about piracy of television programming. The model as practiced at present can't effectively leverage the economic benefits of hyperdistribution, but that model was created before hyperdistribution was technically possible. The age of hyperdistribution demands the development of new economic models which can harness piracy, for profit. So, let's move directly to a discussion of one such model.

Consider Battlestar Galactica. A few weeks before the series premiered on television, I sat down to watch the 13 episodes of the first season, all of which I'd found on BitTorrent. Somewhere around the second or third episode I became briefly aware of the "bug," the smallish, semi-transparent station ID which has become the constant on-screen companion to all television broadcasts. I was looking at the bug for SkyOne, the British satellite broadcaster, which nestled comfortably in the upper left-hand corner of the screen. I noted the bug, then proceeded to ignore it. But it never went away. In episode after episode, the bug remained, a tattoo commemorating the trip from broadcaster to audience.

Somewhere around episode seven, it hit me like a ton of bricks: I was looking at the most valuable and most underutilized piece of real estate in the world. The bug carried the station ID — which is fine if I'm in the UK. But in Australia SkyOne has no meaning at all. So that message, which should be full of meaning — full of "payload" — has been utterly misspent. It's as if they took the finest piece of land in Sydney Harbour, say where the Opera House resides, and decided to use it as depot for broken trains. That screen real estate has real value, because it commands the audience's attention, constantly if subconsciously.

What if, instead of carrying the broadcaster's station ID, the bug contained an advertiser's payload? I decided I wanted to see what that might look like, so I took an episode of Desperate Housewives and ran a little test, using the logo of one of Australia's best known retailers, Myer. I placed the advertiser's bug in the lower left-hand corner. This is probably sufficient for a well-known retailer like Myer (or Macy's or Harrod's, etc.): it's simply enough to remind the public that they exist - and that there's undoubtedly a sale on.

While I thought I was being truly innovative in my thinking, I was wholly wrong. On a recent Friday evening I sat and watched a rugby match: to my astonishment, I found that a commercial broadcasters had already adopted this technique. When the game went into an instant replay, the icon of an Australian liquor distiller Bundaberg Rum did a little dance in the upper left-hand corner of the screen. This means that the technique is already in use, and advertisers understand its value. That's a very important point: advertisers are ready for this.

The earliest models of both commercial radio and television developed around the idea of program sponsorship: one sponsor per program. Over the 1950s (in the case of television) this model evolved toward the 30-second advertisement, which interrupted the broadcast. For the last half-century that has proven to be an enduringly successful economic model, but that model is now under threat from Personal Video Recorders (PVRs), which allow a viewer to fast-forward through all advertisements, often taking them in 30-second leaps, so the audience never sees so much as a single image from an ad. PVRs, playing into the television-taught behaviors of immediacy and convenience, have proven immensely popular, and are not going away; instead, they will become an integral and expected feature of the television viewing experience. This means 30-second ads are not a part of television's future. They're too easy to edit out of the viewing experience.

The idea of an advertising payload attached unobtrusively to the television program has a certain appeal; it can be ignored, but it's always present. The audience can't edit it out of the program without destroying the content of the program. Audiences will learn accept them — so long as the advertisements aren't too busy, distracting, or otherwise obnoxious. (Consequently, there will be a lot of work going on in the next decade to determine just how obnoxious such an ad can be before the audience objects to it.)

As the advertisement-as-interruption disappears, we will see a series of advertisements — perhaps running five minutes apiece — embedded into the programmme itself. This is easy to achieve technically, and will be palatable to most major advertisers. Since this evolution seems inevitable, another question comes immediately to the fore: what's the role of the broadcaster in this new economic value chain? Today the broadcaster aggregates audiences, aggregates advertisers, puts commercials into the program breaks, and makes a lot of money doing this. But — and here is the central point I'm making today — wouldn't it be economically more efficient for the advertiser to work directly with the program's producer to distribute television programming directly to the audience, using hyperdistribution?

Let me run some numbers for you, based on another set of back-of-the-envelope calculations: If we presume that the advertiser is going to pay at least as much as the broadcaster for hyperdistribution rights to a program, there's a large fixed cost for the purchase of those rights. Further, there's another fixed cost to maintain the internet servers which "seed" the program's hyperdistribution - the internet equivalent of broadcast transmitter operation costs. Add in a small amount for the post-production costs incurred to affix the advertiser's payload to the program, and we're done. Those are the entirety of the costs.

The advertiser is looking to lower costs in advertising; if those advertisers are paying between $250,000 and $500,000 for thirty seconds of advertising (in the United States), just a handful of advertisements would cover hyperdistribution costs. It's a numbers game: if enough viewers watch a hyperdistributed television program, it is cheaper for advertisers to work with producers, and handle the distribution themselves. Furthermore, if the program is widely popular, it is far, far cheaper to do so. In other words, the higher your ratings, the cheaper the advertising. That's precisely the reverse of broadcast television, and one big reason that advertisers will find this model so appealing.

Although no formal surveys have been conducted, it's reasonable to assert that at least four percent of Australians, two percent of Britons, and one percent of Americans are already using broadband hyperdistribution to get some percentage of their TV programs. Based on my own research, I have found television downloading to be widespread among men 18 to 25 years old, precisely the demographic most coveted by advertisers. In other words, the prime audience is already there, already waiting and already willing to receive. All that remains is to put the components of this new value chain into operation.

Download (using BitTorrent, of course) the live presentation of "Piracy Is Good?", delivered by Mark Pesce on May 6th, 2005 at the Australian Film Television and Radio School in Sydney. (200MB)

See BitTorrent.com for help downloading.

Continue Reading:
Part Two: The New Laws of Television

Post-Script: The Swarm Manifesto
Written in response to news that the MPAA has filed lawsuits against six sites for sharing TV programs.



Mark Pesce is the co-creator of the Virtual Reality Modeling Language (VRML) - the first 3D interface to the internet - and the founder of the Interactive Media Program at USC's School of Cinema-Television. In 2000, Ballantine Books published Pesce's The Playful World: How Technology is Transforming our Imagination, which explored the world of interactivity through a detailed examination of the Furby, LEGO’s Mindstorms and the Playstation 2. In late 2003, Pesce was invited to the Australian Film Television and Radio School, with a mandate to redesign the curriculum to incorporate the new opportunities offered by interactive media.

Starting in June, Mindjack will be serializing Mark Pesce's new book, hyperpeople.
Sign up to our mailing list to receive reminders about it and all other new Mindjack features.

advertise here
email for info