The End of Showtimes

I planned on taking on this topic eventually, but Netflix's recent announcement brings it to the fore. As distribution technology improves, the limitations that led to mass distribution by synchronized broadcast disappear. Complicated political, social, and technical interactions will shape the transition, but science and science fiction long ago foresaw the shape of equilibrium: universal, on-demand access to content. From Paul Otlet's Mundaneum to Star Trek's ubiquitous and simply named "Computer", visions of universal on-demand access to content predate the earliest television dramas (though the scheduled broadcast slot dates back to radio dramas, of course). While prescience is in limited supply even for recurring visions of the future, content providers such as Netflix and Hulu have shown the strength of the basic idea, and as faster internet access spreads, the logic of broadcast slots will ebb.

The interesting bits, of course, reside in the details, and that part of future visions tends to be dated, or just an opportunity to punt. Sidestepping the more technical questions, the details that I'm planning to become outdated for guessing speak to organization of the production and distribution of television content, and the interaction of that organization with social forces. I make some strong assumptions. One, many people will continue to consume content on devices we would recognize as televisions, even if those devices increasingly provide access to other forms of content. People will still gather around screens, together, and engage in consumption of visual content they do not interact with much (directly). Two, despite declining costs for producing content, demand for productions involving extensive costs and the combined effort of numerous skilled individuals will not go away. Such demand has been around since at least ancient Greece, and reports of its demise remain, so far, greatly exaggerated.

The end of showtimes is the end of simultaneous broadcast as the primary means of television distribution. Consider Time Warner's tentative venture into live television via the iPad. Ignore for a moment the silly restrictions on being in the home, having a Time Warner cable subscription, and being connected using your Time Warner internet connection. Consider the basics of the proposition: live television on the iPad. The idea is compelling, except it contains an extraneous word: live. People who consume television on computers and mobile devices bypassed "live" long ago. In the short term, Time Warner has made what might be an important move (if they tear down some of the silly restrictions; it can't just be a value add for their television and internet service). In the long term, "live" television declines. This doesn't mean that the practice and significance of periodic content releases will disappear, but that showtimes are going to transform into release times. For most content, this will change the viewership graphs from a series of spikes marking the march across the television markets into a smear that still rises and falls to the rhythm of local lives. Some content releases will still be major events, though, a generation's American Idol or Seinfeld or Cheers, with viewership that spikes dramatically on initial release.

The transformation from showtimes to release times affects many dynamics taken for granted in television. The sliced up television schedule, with bands of channels neatly sliced up in time, collapses. Since the disadvantage of releasing too many things in one time slot is lowered by the smearing of consumption, more things are released at times that interact best with perceived viewerships, and fewer are released at less desirable times. Competition changes from eyeballs at a particular time to eyeballs' complete set of television options, driving homogeneity as viewers are forced to choose less among high quality options that previously overlapped in time, and driving heterogeneity as viewers are better able to aggregate shows of interest across a wide variety of venues and release times.

Channels will change dramatically. In one sense, they will no longer exist (this is far future prognostication; their death will be extremely slow), as their shape will be gone. There will be no more bands of non-overlapping time slots. In another sense, they will remain, in whatever form content mediation takes. Some features of channels, such as immediate gratification content recommendation (roughly filled now by reruns), will become increasingly filled by more dynamic, consumer-driven systems. I think what we call networks now will smear together with the idea of channels. There will be some entity buying X number of this show and Y number of the other, and then selling access to the content it manages the creation of.

But channels also have a role on the consumer end of distribution, and how that will play out is very unclear. The most popular streaming providers, Netflix and Hulu, don't really look or act like channels, except mildly with Hulu's advertising sales (and arguably the special Criterion section). More specialized streaming providers, such as Crunchyroll, feel more channel-like, engaging in (somewhat) greater content discrimination and focus. I do not know where the equilibrium will end up. My hope is that the increased costs of cross-device, cross-service (number of devices times number of services is already a heck of a lot) development drive the creation of standards allowing content providers to more easily plug their content into the systems of content distributors. Similar dynamics drive channel activity now, but the restriction on time slots adding to 24 hours stifles growth. Additionally, even now channels tend to acquire content from a parent network or produce the content in house. I believe the result with the time slot constraint relaxed and more free association of content producers with content distributors will be a more scale invariant distribution, where the the size of distributors is inversely proportionate to the number of distributors of that size. In practical terms, this means there will be one or two very large distributors, a few medium sized distributors, and a lot of small distributors. The main thing missing at the moment seems to be the variety of small distributors, though only if we don't count illegal sources, which is not surprising given the high cost of entry. Of course, if my prediction about standards fails, or legal structures appear, or I'm just plain wrong, all could be different.

Now, where does advertising fit in? I believe advertising remains an important part of television funding well into the future. Advertisers pay for access to audience, and the new consumer distribution systems will be able to deliver that audience. Consumers, however, seem less likely to tolerate lengthy ads for streaming shows, meaning there will be fewer ad slots available for the same number of viewers. In compensation, though, ads can be tailored much more directly to individual viewers, a capability which nuances are still being explored, especially for full length television shows. Another direction advertisers are exploring is more complete sponsorship, best demonstrated by Subway and Chuck. Consumers who feel a company shares in the experience and values of a show may be much, much more willing to support the company than typical advertising expenditures. However, such relationships involve risk. Tying a company to a show can end up as an embarrassment to one or the other.

With all the changes that will be coming, these or entirely otherwise, how shows are created will change. I am particularly suspicious of the "pilot" system. Given the changes in how viewership works, with more fluid audiences, rolling out a pilot to test the waters feels a little strange. At the same time, judging the audience for a show will still best be done by creating content and seeing what people like. I have a fondness for the idea that shows will more often be commissioned in small runs sufficiently long to interest actors and amortize fixed costs, but sufficiently short to keep total costs to acceptable amounts given the level of risk involved in a new show: pilot seasons instead of pilot episodes. A pilot season might be somewhere from four to eight episodes. What is more, even if a show does not do as well as expected, it may be possible to continue the show at a reduced budget after the pilot season, if there is an appropriate audience to target it at, since the cost of not running a different show, which is currently mostly the time slot involved, will be much lower. If costs for content creation drop low enough and merchandising becomes sophisticated enough, entire seasons of a show might be commissioned at once, much as already happens with anime in Japan. More and more shows will begin as independent productions picked up by larger content distribution channels and content producers, such as Red vs Blue and The Guild.

In short, technological advances will eliminate showtimes. Their replacement, release times, will drive a slew of changes as content distribution becomes more individualized and accessible, forcing content producers and advertisers to adapt, possibly by forming even closer relationships (insofar as they can draw the consumer into those relationships). The exploration and promotion of new shows will expand with increased individualization, possibly leading to small pilot seasons instead of pilot episodes, and definitely leading to increased uptake of independent productions.