Internet adventures

Since Netflix started putting out its original series by releasing every episode of each season of each show all at once, a move inspired by the phenomenon of people “binge-watching” numerous non-original series on the service they hadn’t originally watched as they came out, there has been debate over what the best strategy is for releasing serialized scripted content in the Internet age. Certainly it would seem that, freed from having to meet the needs of a linear television schedule, there’s no reason not to release content on any schedule you want; most online video series on places like YouTube, both before and after Netflix came along, have been released on a TV-like weekly schedule (or less), but they tend to be made by individuals with low budgets and without the backing of a large company like Netflix, and so need to release episodes pretty much as they’re made in order to maintain revenue to make the next episode. Whether or not Netflix’s strategy is the best strategy, with or without the constraints of a linear network, is another matter entirely.

Certainly, if the series themselves aren’t that important to your business model other than as content to fill out the service, and the main goal is simply to maintain engagement with your product, the binge-release method makes sense; it ensures that you have a large batch of content, presented as a single unit, that people can then consume over an extended period and eventually finish the season without needing to be reminded to come back. On the other hand, Netflix’s main source of revenue comes from its subscription fees, and so what would seem to be best for Netflix’s bottom line would be to keep people subscribed to its service for as long as possible, especially since, like most streaming services, Netflix offers a one-month free trial, thus opening the possibility of people binging an entire season of a given show in a single month and then quitting without paying one cent. (And that’s a very viable proposition; Stranger Things has only eight or nine hour-long episodes in a season, meaning you only need to watch two episodes a week to catch up within a month. It’s entirely viable to catch up in just two nights, or even one day if you can spare the time.) And while Netflix shows are often subject to a burst of intense buzz right around when each season comes out, it quickly dies down as people finish the season and move on to other things, meaning Netflix shows don’t get the same sort of sustained buzz over a period of months as week-by-week shows like Game of Thrones do.

From a creative standpoint, the Netflix model probably does better justice to intricately-plotted shows that in the past might have been deemed better on DVD, where individual episodes don’t necessarily hold up all that well on their own, except in terms of their contribution to the larger narrative of the show, and so their momentum and the immersion in their world is better maintained by watching them in larger chunks. Indeed, since the length of each episode isn’t fixed by the needs of a linear television schedule either, the only criterion for where to place episode breaks at all is to identify good stopping places for people to break at in the likely scenario that they can’t consume the whole season as one really long movie. But this can be a double-edged sword: for truly compelling shows, especially those with lots of plot twists and mysteries inviting speculation as long as they remain unsolved, the week-by-week wait for each episode only strengthens the anticipation. There’s a reason the cliffhanger and other devices borne of the serialized format have such a long and time-honored history. For particularly complex, multilayered shows, the lack of answers drives fans into endless speculation, poring over scenes for clues, rewatching the series in lieu of any new episodes, and generally gaining a deeper appreciation of the series than would be apparent in a one-time surface-level viewing. With a binge-release model where everyone is watching at their own pace, discussion of the show on online forums becomes nearly impossible, with the need to accommodate people at every level of progress through the season. As good as Netflix’s shows may be, they can never truly amount to “water-cooler talk” if not everyone is at the same point.

Of course, the sort of show that creates this sort of constant, edge-of-your-seat anticipation for each episode is also the exact same sort of show that is best suited to a slot on linear television in the Internet age. A show that doesn’t have people feeling they have to watch it the instant it comes out, lest they be left behind in or spoiled by the discussion, is probably also a show that doesn’t lose much by being released all at once and may be better consumed that way, so it’s not clear that there’s a situation where Netflix gets a show that’s better consumed in a serialized format. Still, what this suggests is that the best strategy will ultimately depend on the show. Some shows may work better with a week-to-week release schedule to heighten the anticipation for each episode, others may work better released all at once so they can be consumed as a unit right away, and that’s not even getting into purely episodic shows that would be fine in either format. (I talked about some of the factors going into either strategy, in another context, nearly a decade ago.) It’s not even like you’re bound to one release strategy or the other. You could release episodes in batches, breaking at a point you feel is a good place to leave off and leave the fans wondering, or at any frequency you like that best balances anticipation, attention, and the momentum of regular releases.

Which brings me to, of all entities, Cartoon Network.

Read more

In its highfalutin’ ideals, the Internet is dedicated to the notion of delivering a world of information to all for free, accessible for all to contribute to, available whenever and wherever you want it. Cable television, by contrast, delivers only the content the cable company sees fit to provide you, the vast majority of it from nine companies and laid out on a rigid linear-television schedule, and forcing you to pay for all of it to get just the few programs you want. It is, in short, the antithesis of net neutrality, and as I laid out earlier, the cord-cutting debate is effectively a clash between the outdated vision for the provision of content and the new vision sweeping it aside.

It is therefore tempting to conclude that traditional, wired linear television has not only outlived its usefulness but needs to be actively destroyed to preserve the ideal of net neutrality, that any future it might have is only in schemes such as Comcast’s “Stream TV” that use the existence of linear television over traditional wires to circumvent net neutrality by providing preferential treatment to certain types of video content. But as I’ve chronicled, at least in the wireless context, linear television can not only play a key role in network management, traditional broadcast linear television can actually benefit net neutrality by allowing those tools to be accessible to as many potential programmers as possible and delivering it to everyone regardless of carrier.

Not all of these benefits can be directly transferred to the wired context. Most obviously, wired connections are inherently dominated by whatever company owns the wires and processes the connection; they are inherently going to have some control over what content can use the linear television wires. And if you believe that at some point, everyone is going to be fed by a direct fiber-optic connection, fiber-optics sees no benefits to linear television at all; each customer has its own fiber leading directly to the node and the rest of the system without being impacted by anyone else’s activity. Unless, that is, it’s set up in something like a passive optical network, which reduces the cost and amount of fiber needed to go into business by splitting one fiber to serve multiple endpoints, effectively delivering each customer’s content to every customer on the same splitter. It requires end-terminals to sort out what content belongs to which customer and encrypting content to prevent eavesdropping, but if I were running such a network I would look into a way to actually foster eavesdropping for live streaming video, so if someone is watching a game on ESPN3, anyone else on the same splitter that wants to watch the same game can ride on the first person’s stream, freeing up network capacity for everyone else.

This hints at what the future of wired linear television might look like on both coaxial cable and fiber-optic networks, blurring the lines between online and linear content so thoroughly it may be impossible for the end user to tell which is which; some WWE Network content might be delivered linearly, while in some areas a network like Logo might be delivered via IP. It doesn’t have to look like Stream TV, but it might have to adopt some of the less savory elements of T-Mobile’s Binge On, finding some way to identify live streams so they can be isolated and delivered once. It may also be beyond the capability of some providers to pull this off dynamically, certainly while minimizing lag; linear delivery may be something that has to be arranged ahead of time in some way. The key, then, is to figure out how to foster such a system while preserving the principle of net neutrality.

Last year I sent a Congressional committee 13 pages of comments explaining what principles I felt should underlie any revisions to the Communications Act to update its treatment of video for the Internet age. More recently my thoughts have coalesced with regards to how to regulate wired television to bring it into alignment and consistency with the principle of net neutrality and existing rules governing the Internet. To that end, here’s what it might look like:

• No content delivered via linear television, particularly that not delivered over-the-air, can be withheld from delivery over the Internet, whether by the ISP or the content provider. This is a basic way to avoid using linear television as an end-around around net neutrality rules, and it also demonstrates an important difference between linear television now and going forward: it is no longer a prerequisite for the delivery of content. ISPs may no longer find it necessary to carry a linear television service at all. Really, the notion that cable operators are just a backdrop and infrastructure system through which programming is carried, which has little to no role in what programming it is, is one that probably should have taken hold several years ago as digital cable obviated the condition of scarcity that had governed the pay-TV industry, and certainly once Internet streaming became competitive; it’s still impossible for a cable operator to carry every would-be linear network, but the notion that your ability to watch a channel anyone would actually watch would be determined by what pay-TV operator you subscribe to seems not only quaint and outdated, but antithetical to net neutrality. In this vein:
• If linear television is carried over any wired network, it should be controlled at the infrastructural level. I’ve seen suggestions to increase competition for Internet service by decoupling control of the infrastructure from the service delivered over the pipes, recognizing the natural monopoly the pipes themselves represent while still allowing ISPs to compete freely (i.e., without utility regulation) for the service delivered over them. If so, the same principle underlying broadcast television applies: the content allowed to benefit from linear television should be the same for everyone regardless of carrier. Any distinctions between the content available in different areas should be a natural result of the existence of different networks, without individual ISPs controlling what content you can and can’t watch and how easily you can do so.
• Keep the existing must-carry rules. Ideally, broadcast television serves as a means to deliver the most popular content quickly and efficiently to a variety of devices. It’s reasonable to think that the same content would also be most popular over a wired connection. This rule would establish some degree of parity between wired and wireless video, ensuring the same content that benefits from linear television for the one gets the same benefit on the network with more capacity, and recognizes the original purpose of cable TV to deliver broadcast stations to areas their signals couldn’t reach. The increase in capacity may mean it’s actually not necessary to add any further linear channels beyond those provided by broadcast stations, but it’s reasonable to think video consumption would be higher, and be at higher quality, over a wired connection. If there is a need for further linear channels:
• Forbid any monetary transactions as a condition of linear carriage between an ISP (or infrastructure authority) and a content provider, especially per-subscriber or per-viewer fees. Content providers may only charge consumers directly, though they may use any scheme they wish to do so as long as the ISP (or infrastructure authority) is not directly involved. This rule ensures that potential network strain is the only factor going into what channels are carried linearly, and should render retransmission consent unnecessary and contradictory. Net-neutrality foes and general free-market advocates may look at this rule and think it’s cutting off “innovative” business models, but because in this model linear television is a means for smoothing over the transmission of content that is being made available through the Internet anyway, one that should reduce the costs to both parties relative to the alternative as the number of viewers goes up, there should be no relationship between the costs of maintaining a linear feed and the number of subscribers an ISP has that would have access to the feed or the number of people using it; the costs of providing the feed should be the same regardless of how many people are on the network or can view the feed, indeed that’s entirely the point of linear television in the 21st century. The cost of providing and acquiring the content (an issue concerning the consumer) should be a separate issue from the cost of delivering it (an issue concerning the content provider and ISP). There is therefore no good reason to charge ISPs anything other than a flat fee (and even that’s questionable; small ISPs wouldn’t be too hurt by being unable to amortize the cost across more customers because they’d have less need for linear TV to begin with, but any sort of payment scheme might create the same imbalance that made retransmission consent necessary), and for an ISP to charge a content provider would amount to exactly the sort of “paid prioritization” net neutrality advocates fear so much.
• Impose effective protections against discrimination and an effective dispute resolution process. Again, the only factor that should go into whether any programming is or is not offered linearly is its raw popularity, and any content popular enough to warrant it should have the opportunity to use it regardless of their level of resources or connections. This does not mean disadvantaged groups or public entities should have a blanket right to a linear stream; again, a linear channel is not a prerequisite for the delivery of content, and in some cases what they’re looking for no longer even needs to be video. In many cases these entities seek to take advantage of the modern cable-bundle model to acquire production and distribution resources on par with more well-heeled groups they might not be able to attain if they were forced to stand and fall on their own merits. It may be desirable to introduce new programs to benefit minorities and open up the possibility of using franchise fees on ISPs to fund the production of video and other content by public entities (similar to today’s “public, educational, and governmental channel” system), but linear television is irrelevant to that discussion and shouldn’t be hijacked to attain those goals.
• The quality of a linear stream must be the same across all platforms and an ISP (or infrastructure authority) must not degrade it. Because an ISP does not have to carry a channel (except for a broadcast station) for its customers to have access to the programming on it, there is no reason for an ISP to reduce the quality of a stream in order to fit in more streams. If a content provider offers its stream at too high a quality and causes all ISPs to balk at carrying it, that’s a market signal to reduce the quality of the linear stream. There is one exception to this rule: because the main purpose of over-the-air broadcasting is to reach mobile devices, it may be beneficial for broadcasters to offer wired services their content at a higher quality than they broadcast it over the air.
• Consider imposing restrictions on how many streams one entity can control 24/7 (or otherwise beyond particular events) on one set of wires. Ideally, impose some degree of parity in ownership restrictions between broadcast and wired linear networks. In particular, ISPs should be severely restricted in how many linear streams over their own wires they directly control the content of. In general, consider moving to the Canadian system where the FCC has as much latitude to regulate cable networks as they do broadcast networks.

These rules provide a baseline to move wired linear television away from being a marketplace defined by 1990s rules and market realities and towards becoming a tool that enables the benefits of both linear television and the Internet to be available to all, consistent with the ideals of net neutrality. They hold the potential to usher in the dawn of a new era of consumer choice that frees Americans from the paradigm of having one’s entertainment options dictated by the cable company and enables them to choose from a menu of options that provide value to them. It’s my hope that this provides a framework for policymakers to rethink the wired television landscape and for the American public to imagine what it might look like.

An important step in the dissolution of the distinction between linear television and online video content was taken (or at least started) this week when FCC Chairman Tom Wheeler called for the introduction of a new standard to make cable TV content available to devices other than the cable company-provided boxes most people pay handsome monthly rental fees for, allowing people to watch both pay TV content and streaming video on a single box.

Immediately, the cable industry tore into Wheeler’s proposal, forming something called the Future of TV Coalition to be the public front to its opposition. Most of the critiques I’ve seen of their criticisms don’t seem to go much further than just dismissing it as whining about potentially losing their lucrative set-top-box rental fees, but that may be because it comes off as completely unhinged, seemingly attacking a completely different plan than what consumer groups and Wheeler are advocating and making points the proposal has already addressed or that make no sense, a strawman bearing no resemblance to what’s actually on the table. They claim the proposal would allow tech companies to cut up and resell pay TV content as their own, but it makes no sense and there’s no reason to believe that people wouldn’t still pay their cable company to deliver content, only having a choice in what device would deliver that content. They claim the proposal would allow tech companies to muck with channel ordering and numbering in violation of contractural agreements, when it would be trivial to require any interface to leave the channel lineup alone (or at best to leave any reordering to the consumer). They claim the proposal would mandate the installation of an additional box when the proposal specifically advocates a software-based solution, the whole idea of which is to allow the provision of pay-TV content on boxes that already exist (including cable companies’ own boxes) or on no box at all. They claim the proposal would strip out security and privacy protections when the whole point of it is to arrive at some sort of solution to deliver those protections and credentials to independent boxes for them to process just as today’s cable boxes do.

Though the cable industry has publicly supported past and present efforts to open up the set-top box market, and claims to be all for opening up access to a wide variety of devices, they spend a lot more time bashing the FCC’s proposal than suggesting their own alternative, despite claiming not to know exactly what the FCC’s proposal is. Instead, the closest they come to suggesting an alternative is to repeat the word “apps” over and over. The commission’s proposal, they claim, is unnecessary because Tim Cook says “the future of TV is apps” reflecting the “apps revolution” of consumers, programmers, and cable companies embracing the “apps-based model” making “apps” available to millions of devices and apps apps apps apps apps.

If it seems odd that consumer groups, tech companies, and the FCC would be against television delivered through apps, given the entire backdrop under which this whole thing is taking place, that’s because they aren’t; indeed, the proposal specifically names “app developers” as being among those that would have access to the necessary data to effectively and securely deliver pay-TV content. So it’s not clear what the cable industry means by the “app-based approach” that wouldn’t include the very concept they’re contrasting it to. What they seem to be trying to say is that the FCC doesn’t need to do anything at all, because the existing apps available already deliver pay-TV content to the sorts of devices that would benefit under the proposal (not that that would keep them from pushing expensive cable box rentals on people) – though it’s not clear what kind of apps they mean, because they alternately cite both TV Everywhere apps provided by programmers as well as cable companies’ own apps, all in order to support letting the market do its work rather than an actual, concrete proposal. Certainly what they say sounds reasonable enough at first glance, but what sort of “apps-based approach” do they actually advocate, and what exactly is the problem consumer groups and the FCC have with it?

Well, according to the report released by the FCC’s Downloadable Security Committee, cable companies’ proposal would involve delivering content to devices using an app and user interface provided by the cable company, leaving consumer groups concerned about precluding other entities from innovating with their own user interfaces. Of course, forcing people to use a single app to access all linear cable TV content is precisely the opposite of what the “apps revolution” is actually about: decentralizing access to content and allowing content providers to offer their wares to consumers directly, with an experience they can control themselves.

The cable industry seems to believe, or wants the FCC or public to believe, that cable companies’ services themselves are the product, rather than the content offered through those services, even though that content is mostly the same from one cable company to another. This belief is betrayed in their listing of the top “video subscription services”, which lists the streaming services Netflix and Hulu alongside the top cable operators and satellite providers, as though Netflix and Hulu’s primary competition were cable companies themselves, not, as Netflix itself has identified, content providers like HBO. Perhaps cable companies’ greatest fear isn’t losing the billions of dollars in set-top box rental fees, but that in divorcing them of that the FCC might recognize that the real “future of TV” is one dominated and identified by content providers, with cable companies merely providing the backbone through which that content is delivered, and that they might accelerate that future by providing the tools for their wares to be offered through an experience completely divorced from the cable companies’ control. To be sure, content providers might feed this misconception; the contractural concerns such as channel placement cable companies worry about the FCC’s proposal undermining are rooted in a notion of a single lineup of numerical channels defined by the cable company, and perhaps a proposal that makes it irrelevant is one that should be considered and adopted. Content providers would no longer be able to get cable companies to try to force-feed their content by placing it near content that’s actually popular, but even their own TV Everywhere apps (which would seem to have little reason to exist on smart TVs and devices like Roku as it stands) would stand to benefit by being able to access the cable company’s linear feeds directly, strengthening their own brand by making it easier for those apps to become the primary gateway to their content.

The cable industry is right that the “apps revolution” is changing the way we watch TV. The reason they’re opposing the FCC’s proposal so strenuously is that they know it holds the potential to make it all the more successful at it.