Portals: A Treatise on Internet-Distributed TelevisionSkip other details (including permanent urls, DOI, citation information)
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License. Please contact email@example.com to use this work in a way not covered by the license. :
For more information, read Michigan Publishing's access and usage policy.
1 Theorizing the Nonlinear Distinction of Internet-Distributed Television
The key distinction of internet-distributed television from that of broadcast or cable distribution is that it does not require time-specific viewing. Although distribution technologies using internet protocol can also deliver content in live and/or linear fashion, this has not been its dominant application to date. It is by eliminating the necessity of time-specific viewing that internet distribution allows different logics that extricate it from many of the conventions that have been established for television industries built on linear distribution.
The full gamut of legacy television industries’ production and distribution practices, particularly those of scripted series, assumes time specificity. For decades, audiences were forced to organize their viewing according to a network-mandated schedule. Almost all the conventions of television—a flow of content, program length, expectations of weekly episodes—derive from practices developed to cope with the necessity of the linear schedule. Internet distribution advances what VCR and DVR recording and DVD access began to allow so that viewers can select television viewing just as they might select a book from a library.
In addition to the way time and timeliness has defined a particular viewing experience, time also has played a key role in circumscribing legacy business practices. The business of scripted series created for broadcast and cable distribution relies on time-based windowing of content spread over a series of markets (first, second, etc.) and licensing practices that have imbedded “time” in the core economic transactions of linear television. The network licensing the original airing of a series typically enjoyed exclusive access to the series for an initial period. Much of the economic exchange between studios and networks was predicated on exclusivity that could be enforced by the scarcity characteristic of linear distribution that severely limited access to programming in an era before DVDs and internet distribution.
Linear television’s time constraints also supported the advertiser-reliant business model that dominated the network era. The distribution bottleneck that yielded limited availability of content aided a system built on creating entertainment to gather an audience that would then watch advertisers’ messages. Of course, advertiser support was not required by this distribution technology—as robust public service television marketplaces elsewhere illustrate, but it was the American experience.
Importantly, a lot of television programming cannot be easily removed from a schedule. Much of its utility or value is connected to its liveness or correspondence to rituals of daily life, and linear broadcast and cable distribution remain effective for such content. Internet distribution consequently might not be expected to wholly replace previous forms of distribution. Rather the core of its disruption will be on types of television markedly improved by the nonlinear affordance of internet distribution. This leads to a focus in the following pages on scripted series as the content form for which new logics of internet distribution are most profound.
Paradoxically, the current developments that seem revolutionary—be they internet protocol technology and related affordances, strategies related to a previously peripheral revenue model, or changing viewer behaviors that emerge from these technological and industrial shifts—are not as radical a break as has been commonly presumed. Antecedents for the contemporary environment exist as far back as the circulating libraries of the 1700s and more recently in the video rental businesses of the 1980s through 2000s. Internet distribution of television and the shifts in practices that derive from its affordances allow behaviors that were peripheral in an age of analog, physical media such as time shifting, self-curation, and á la carte access to become central and industrialized practices.
The need to explain this process of shifting cultural uses of technology brings to mind Raymond Williams’ distinctions of emergent, dominant, and residual technological practices, but this framework asserts too strong a notion of replacement than evident in this case. Rather, at this, admittedly preliminary point, the relationship between technologies of distribution is better described by what William Uricchio describes as profound “pluriformity” although, in time, more evidence for Williams’ pattern may emerge.
Theorizing Nonlinear Television
In his 1989 book, The Capitalization of Cultural Production, French sociologist and cultural industries theorist Bernard Miége organized the logics of media production into three models: the publishing model, the flow model, and the written press model. In organizing wide-ranging media industries into these models based on their general characteristics, central function, economic organization, creative professions, business models, and market characteristics, he established “logics” useful in analysis and theory building about media industries that could be broader than single industries but still narrow enough to allow useful theory development. Within Miége’s tradition of Francophone cultural industries analysis, scholars identify the logics of media industries and discern patterns of behavior, as well as possibilities foreclosed by the difficulty of deviating from the established logics, to develop critical understandings of these industries and the creative possibilities they encourage.
Over the last three decades, Francophone scholars of cultural industries have defined and reconceptualized “models” through which cultural industries generally can be categorized in order to make claims of the norms and operations of media sectors and to engage in theory building broader than a particular case. Miége’s triumvirate of models—publishing, flow, and written press—circulated most widely. Various refinements to the models were subsequently proposed; some suggested that the publishing model might be better classified as “commodity” production and that the dynamics of written press and flow industries were like enough to be combined. Others, including Patrice Flichy, have offered different conceptualizations of larger-scale “models” relative to particular practices or “logics.” Jean-Guy Lacroix and Gaëtan Tremblay expanded the possible categorizations by proposing the “club logic” (really a model in Miége’s use of the term) in response to the conditions described here as a subscriber model of generating revenue from media goods.
The tools of cultural industries research are ideally suited for investigating the emergence of internet distribution of US television, as it is clear that much of what was known of the business of television is inadequate for understanding these new distribution technologies. The affordances of internet distribution—particularly the possibility of nonlinear distribution—establish different logics that in turn allow distinctive protocols that enable an array of practices and require an expansion of Miége’s tripartite organization of models.
Miége identified the “flow model” as characteristic of radio and television and what were emerging in the late 1980s as “new media.” The flow model is primarily distinguished by the continuous flow of the goods produced and the way these media correspondingly become integrated into patterns of daily life. The scarcity constitutive of broadcast technology’s capabilities required the distribution of a daily schedule of programming and led to its operation within the flow model. The main activity of industries operating within the flow model is as program planners who schedule the flow. Flow industries create—or arrange for the creation of—the components of the schedule (shows), but Miége argues flow industries primarily produce a schedule rather than particular creative goods. Broadcast and cable television distribution unquestionably followed the norms of the flow model that Miége set forth. The break from linear flow that internet distribution enables thus requires a different model for understanding nonlinear television.
Internet technologies dislodge television from the characteristics of the flow model and its attendant industrial practices so much so that some forms of television—particularly scripted series increasingly engaged through self-selection from on demand access to portal libraries—require a different model of conceptualization. Miége’s “publishing model” provides another option although even this does not provide a fully explanatory model. Media that are “cultural commodities composed of isolated individual works” such as films, books, and albums characterize the “publishing model.” The publishers’ primary duty is to organize the production and reproduction of the good, typically choosing which works will be created and identifying the creative team that will produce it.
Miége’s publishing model encompassed the norms of industries such as film, the recording industry, and book publishing. The publishing model is thus not about media based upon reproducing the written word, as might be implied from the moniker, but of industries built on the creation of cultural commodities that are single works that consequently require extensive marketing. Unlike the flow and written press industries that are defined by the continuity, regularity, and the ritual of consuming their goods, a series of distinct purchases characterizes the publishing model. This attribute alters the logics of media production in a way that encourages other industrial practices. In particular, a revenue model of direct transaction payment for the media good—buying a book or album—or access to view it in the case of theatrical viewing of film dominated such industries in their analog era. Media produced within the publishing model have not substantively relied upon advertiser support.
Nonlinear distribution of television requires rethinking many assumptions of television and recalibrating models and frameworks for understanding the business of television. The practices available to internet-distributed television are distinct from those of the flow and publishing models of media operation. The protocols of portals consequently warrant deeper investigation and theory building to establish frameworks that account for their distinctions and deviations from established norms.
The affordances of internet distribution allow variant logics to govern this form of television distribution, that in turn, encourage protocols and strategies distinct from broadcast and cable distribution that have extensive implications for internet-distributed television. Shifts in industry logics—even of a less substantial nature than switching between models—adjust creative and textual possibilities to such an extent that changes in cultural goods should be expected. The emerging strategies vary television’s role in culture, the creative goods television industries are most likely to produce, as well as the agency and opportunities afforded to creative talent. Only a nuanced examination reveals the points of connection and differentiation among television distributed through different technologies. Exploring the changing landscape of television as not merely a technological evolution but also with attention to shifts in logics and related practices offers a road map helpful for considering the wide-ranging disruptions digitization and internet distribution have brought to other media as well.
Nonlinear Television as Characteristic of the Publishing Model?
If the nonlinearity of internet-distributed television makes the flow model a poor fit to explain its characteristics, might the publishing model be more valuable? Some evidence can be identified that suggests Miége’s publishing model could inform internet-distributed television. To be clear at the outset, there are also several limitations to understanding internet-distributed television within the publishing model, particularly that it is based on transaction payment for a particular good. In most cases, access to a library of content is transacted in the exchange between portals and subscribers, which is a remarkably different exchange.
Nevertheless, thinking about television series within the publishing model is a paradigm disrupting thought experiment that enables richer conceptualization of the consequences of internet distribution. It aids in anticipating emerging business models and practices to allow scholars concerned with critical aspects of media within society to foresee the implications of changing industrial practices related to internet-distributed video for creatives, texts, and audiences. It also strengthens the case for creating a new model better attuned to the specificities of internet-distributed television and its emerging protocols.
What would television organized according to the publishing model look like? I return to this thought experiment in the treatise’s conclusion, but it likely most closely follows the experience of the transaction of books. The brief television-on-DVD market that emerged in the early 2000s also hinted at this experience. Of course, television series continue to be sold on DVD, but both the monetary cost and the inconvenience of acquiring a physical good make this a preferred mechanism of distribution in few cases. The internet-distributed transaction market also remains comparatively small—22 percent of the digital video marketplace in 2014, and forecast to decline in share to subscriber-funded services. Although perfectly suitable as an experience, the monetary cost of transaction as a revenue model has primarily made it an option only for those for whom convenience outweighs price because competing portals offer better monetary value. Notably, with the exception of Louis C.K.’s 2016 Horace and Pete experiment, no series has been created for a first window of transaction sale, which makes transaction an uncertain revenue model for content creation.
Per series transaction conforms well to the publishing model, but portals that offer access to a library of content have dominated the preliminary years of internet-distributed television. Unlike the simple transaction characteristic of the publishing model in which a fee is exchanged for a single good, the portals bundle access to a range of series and movies for a monthly subscription fee. It is ironic, given the attention to “unbundling” and disaggregation as implications of internet distribution in media industries such as newspapers and music, that the most pervasive form of internet-distributed television has relied on bundling films and television series. This strategy has been explored in economics as a practice of tying, bundling, or as a “club model” of media industry operation although none of these contexts precisely reproduces the peculiarities of the transaction of internet portals. Yannis Bakos and Erik Brynjolfsson model the economics of bundling goods sold by internet to explain the value of the strategy. They identify that a seller can more precisely predict how a consumer will value a collection of goods than can a seller value individual goods. The transaction of a library as opposed to single goods is consequently valuable to both sellers and viewers.
The portals dominating internet-distributed television in the United States in 2017 do not conform to the logics of either the publishing or flow models of media production. Portals are not effectively conceptualized according to the flow model because most operate without a linear schedule. Although portals still require tasks of content selection and organization—discussed here as curation—the comparative lack of constraint that allows more than a single text to be available at a time significantly adjusts the logics and strategies of portal curation from those of broadcasting and cable. Even those portals that maintain some linear, flow tendencies—Watch ESPN, Twitch, and the autoplay experience of YouTube—provide users with such an abundance of simultaneous viewing opportunities that the strategies associated with managing the scarcity of scheduling prove suboptimal.
Moreover, in contrast to the norms of the publishing model, the underlying transaction of a portal is access to a set of goods, not the exchange of a specific good. Portals that rely on a subscription fee for access to a library of content thus diverge significantly from the publishing model’s basis in transaction sale. This substantially differentiates the operation of most portals from the logics of the publishing model as well.
How Does Nonlinear Curation Differ from Linear Scheduling?
It is difficult to extricate the affordance of nonlinearity from the protocols it enables that in turn adjust the business of television production and distribution. But the extensive implications of nonlinearity can be seen if we compare a media industry that has never been organized by linear dictates. Juxtapose the practices of an industry such as book publishing—which has no technological capacity constraint—with linear television. Such a comparison reveals how broadcaster’s capacity constraint of only making one television show available at a time deeply structures the parameters of linear television. A single channel can only distribute 24 hours of programming a day. That is a significant limit to what can be “on” in any day. Thus, the channel’s ability, or requirement, to select the one thing available at any time very much defines linear television.
Linear television is consequently characterized by two related attributes: capacity constraint (limited content available) and time specificity (content available at a particular time). These attributes encouraged protocols of television experience, such as the common tendency for people to “watch television,” rather than a particular show, or to sit down to “see what was on.” Such behaviors and expectations derived from the sense of limited availability long perceived as inherent to television. These once-common phrases about television behavior reveal how audiences understood and experienced television, but lack pertinence in a nonlinear environment.
The attributes of linear television necessitated that the guiding determination in schedule construction for advertiser-supported television be selecting the content believed likely to attract the largest audience. Decades of familiarity with this strategy have made it seem natural, but content could have been prioritized for several other reasons: it was the best developed, it brought the most underrepresented voice, or it indicated the highest aspirations for the conventions of the medium. Even once niche address became common among precisely targeted cable channels, the primary strategy remained selecting the content likely to attract the most audience members—just perhaps the most women, children, or sports fans depending on the target of the channel.
Without scarcity governing programming strategy, it becomes possible for other tactics to emerge. Nonlinear distribution eliminates time specificity and greatly reduces capacity constraint. Of course, capacity is not limitless for nonlinear providers. But rather than distribution being the source of confinement, the cost of acquiring content becomes the key limitation. Technologically, a portal could conceivably make any piece of content ever made available, but based on current business models, the cost of licensing such a vast library would be prohibitive. Given this constraint, what strategies govern portals’ selection of content?
At this preliminary point, evidence of two curation tactics can be identified for portals distributing legacy-style television. One is an audience strategy, and the other is a content acquisition strategy. The audience strategy is simply that of curating content to meet the needs of a specific audience or audience taste, especially niches not well served by existing television. For example, Noggin provides a portal with programs for preschoolers; WWE Network features programming interesting to wrestling fans. Although seemingly obvious, this is markedly different from the broadcast audience strategy of creating programming likely to gather as many people as possible.
In some ways, the strategies of audience targeting—or channel branding—that have been characteristic of cable channels seem consistent here and applicable to the portal environment. In some cases clear parallels can be identified, but the nonlinearity of portals enables a deeper deployment of this strategy. Channel branding was valuable in an expanded programming environment to allow viewers to know what they could expect to be “on” a particular channel. The nonlinearity of portals allows much more depth and the development of a library—rather than schedule—to service that interest. Moreover, since the majority of portals rely on subscriber funding, the difference in revenue model requires portals to truly serve their audience niche. In practice, this is the difference in imperative of ad-supported Nickelodeon, which seeks to attract children to sell to advertisers, and the portal Noggin, which seeks a monthly subscription fee from preschoolers’ parents in exchange for access to ad-free preschooler content. Nickelodeon’s brand announces it as a destination for those seeking children’s programming, whereas Noggin must provide enough value to warrant the subscriber fee.
Assessing the value of cultural goods is complicated and requires more extensive theory building. One effort explores how a total value of cultural goods could be comprised of multiple value measures such as nonmonetary returns, market and nonmarket use value, option value, non-use values such as existence and bequest value, and instrumental values. A particularly challenging aspect of theorizing value for cultural goods is that viewers have “bounded rationality,” which means they do not know their own preference for cultural goods. The degree to which viewers often do not know what they want to watch explains the value of libraries that bundle multiple series.
Vastness of library is thus a key attribute for portals but must be weighed against content acquisition costs. An advertiser-funded portal seeks to include as much content as possible to attract the most viewers that can then be shown advertisements. In a transaction model where viewers buy or rent particular goods, a retailer also seeks to offer as vast a catalog as possible because earnings are tied to the number of goods sold. The requirement that subscriber-funded services provide something viewers value enough to pay for makes it likely that the linear niche targeting strategy will not be precisely replicated in nonlinear competition. Exclusivity becomes very important; to earn monthly payment, subscriber-funded portals need to provide content viewers want to watch—rather than just something to watch.
The more complicated site for parsing this strategy is within a portal with a broader array of content, such as Netflix. Extending the concept of cable channel branding, what is the “brand” of Netflix? Netflix’s library contains content targeting multiple taste groups, and it is able to effectively target these multiple tastes because it is nonlinear. Even the most loyal Netflix consumer accesses a small amount of the library and likely has little awareness of what else is available. My description of the Netflix brand based on my viewing might differ markedly from an equally devoted subscriber with different tastes who also finds her needs met by content completely separate from what I view. Netflix takes advantage of what might be considered as the positive properties of filter bubbles so that people with different tastes have very different experiences of the content available in a way that affirms their sense of the Netflix brand.
Varying from the niche strategy of several portals then, Netflix pursues a “conglomerated niche” strategy. The company services multiple audiences, but this is very different than a “mass” strategy. It does not license or develop a series with the expectation that all Netflix viewers will value it, but develops offerings with distinct segments of subscribers in mind. Such a mass customization strategy is made possible by the elimination of the time specificity and capacity constraint of linearity that prevent channels from effectively targeting multiple audiences to achieve scale. The network era comparison for Netflix is not a channel, it is a conglomerate; Netflix is not like Nickelodeon, it is like Viacom.
Such a conglomerated niche strategy achieves the advantages of scale while servicing heterogeneous tastes. What are these tastes? Only Netflix executives may know. Two of the niches targeted by Amazon Video are “people who go to Comic-Con” and “people who listen to NPR (public radio).” And people with kids are clearly another target. Importantly, this is not the same as a mass strategy. Netflix achieves scale that creates efficiency for its operation, but not by being one thing to all subscribers.
Importantly, it is the attributes of nonlinearity that enable Netflix and others using a conglomerated niche strategy to be different things to different people—without anyone really noticing—and make this strategy effective. Netflix’s curation strategy is guided by mass customization and its ability to market and recommend to subscribers is highly individualized. In an interview with television critic Alan Sepinwall, Netflix’s chief content officer, Ted Sarandos, illustrated this logic in action by explaining how Netflix does not blanket the opening screens of all subscribers with each new show because the worst outcome is having the wrong person see new content and thus produce negative word of mouth. This restraint, of only marketing a show to those likely to like it, illustrates an approach wholly uncharacteristic of the mass strategies of broadcast television, but viable and effective for nonlinear distribution. Of course, Netflix is also able to do this because of the proprietary data it collects that inform it of subscriber preferences—another affordance of internet distribution.
One subgroup of Netflix programming receives a lot of buzz and attention because it is one that closely matches that of cultural and taste opinion leaders such as critics. But evidence of multiple constituencies occasionally sneaks through. In a rare moment of data sharing, Netflix CEO Reed Hastings revealed that more subscribers watched its original horror series Hemlock Grove in its first days on the service than viewed the much buzzed about House of Cards. Likewise, the announcement of a multi-film deal with Adam Sandler drew stunned silence from most who faun over the programming that services Netflix’s version of the “NPR” audience. Yet, in an August 2016 interview, Sarandos noted that the Sandler movies had premiered at number one in every Netflix market and that the Sandler film, The Do-Over, was still in the top ten in nearly all markets three months after release. Such data—if valid—reveal the conglomerated niche strategy in operation.
Of course, a key component of the portal experience derives from the mechanisms available for recommending content and helping viewers find programming that matches their interests. The portal experience is constructed through what Ted Striphas identifies as “algorithmic culture.” Similarly, Jeremy Wade Morris identifies the emergence of “infomediaries”—the companies responsible for shaping the cultural goods audiences encounter—that play an increasingly important role in the functioning of digitally distributed media. The actions of algorithms and infomediaries allow mass customization—despite Netflix’s sizable scale—to provide a seemingly individualized experience on the part of viewers.
Portals derive the most value from content that is exclusive to their libraries. Exclusivity—although long a tool of television scheduling—achieves even greater strategic use in the context of subscriber-funded portals. Unlike the norms of linear television in which a channel typically only secured exclusive rights to content for a period of time, subscriber-funded portals have sought to license original series in a way that effectively forces those who desire this content to subscribe to their services. This breaks from the prevalence of price discrimination strategies used in linear television that extracted higher economic value from audiences seeking immediate or earlier access to a series. Such strategies are less useful in a technological environment that makes unauthorized content readily available and diminishes industry-created artificial scarcity.
Also relevant to this strategy is the way nonlinear portals have tools that allow them to maximize niche strategies. As information scientists Michael D. Smith and Rahul Telang explain:
These processes—on display at Amazon and Netflix—rely on selection (building an integrated platform that allows consumers to access a wide variety of content) and satisfaction (using data, recommendation engines, and peer reviews to help customers sift through the wide selection to discover exactly the sort of products they want to consume when they want to consume them)…. They can do this because shelf space and promotion capacity are no longer scare resources. The resources that are scarce in this model, and the resources that companies have to compete for, are fundamentally different resources: consumers’ attention and knowledge of their preferences.
The affordance of nonlinearity adjusts viewer experience, but the affordance of enabling data collection also adjusts the strategies portals can use in developing and curating their libraries.
Entirely separate from this audience strategy for portal curation, portals use a content acquisition strategy that leverages self-owned content to manage constraint in their program budget. Since the main limitation on a portal’s library is the funds it has available to license or create programs, it is not surprising that many of the most successful portals to date are those that have been able to launch with ample content provided by corporate owners that own most or all of the content on offer. CBS All Access provides the best illustration of what I explore as a “studio portal” in Chapter Three. All of the content available on CBS All Access is owned by CBS Productions. Although most of the marketing has focused on the service as a way to watch current CBS programs, the portal also includes a significant number of library titles. In distributing through a self-owned portal, CBS seeks to profit directly from this library, rather than licensing it to another service. Notably, it also has the advantage of gaining considerable insight into the value of its content through the data generated by viewers’ use of the service.
Leveraging self-owned content need not exclusively determine the content of a portal. The comedy portal SeeSo not only draws much of its content from the library of its parent company NBCUniversal, but also licenses content from others to better define its brand and provide value to its subscribers. Even the ability to leverage some content from a co-owned studio provides valuable flexibility for a portal.
Of course, vertical integration is not a strategy particular to portals; it had become a crucial strategy for linear television by the early 2000s as well. It arguably takes on heightened value for library construction because without capacity constraints, a portal’s value is tied to the depth of its on-brand content. Leveraging self-owned content can help a portal offer a value proposition at launch adequate to recruit subscribers, whose fees over time can finance the expansion of that library.
Nonlinear television also provides value to audiences through convenience that may alone be more significant than any content strategy. Research on why viewers subscribe to Netflix found that 82 percent do so because of the “convenience of on-demand streaming programming,” 67 percent because it is “cost-effective,” and 54 percent because of its “broad streaming content library.” In other words, content is only the third most compelling reason for subscription. Filling out the motivations, 50 percent subscribe because of multi-device functionality, 37 percent for kids programming, and 23 percent for original programming. Such data illustrate the value inherent to nonlinear distribution.
Nonlinearity also creates new challenges. The metrics of success in linear distribution were very clear: how many watched when it aired. The value of a piece of content in a portal library is not determined so quickly. Although attempts to assess Netflix’s original programs have sought counts of views within the first month of availability, measurements based on immediacy offer only a partial assessment of content value. It has long been the case that television content operates on what economist Richard Caves describes as the “ars longa” property, or a long arc of revenue return that makes television revenues durable. The tremendous earnings of television series—both in the linear-only past and now—are often achieved decades after episodes are produced. Just as shows sold in syndication such as Friends and Seinfeld continue to produce considerable revenue for the studios owning them even twenty years after they ceased production, so too must the value of content held in portal libraries in perpetuity be measured over the full period of its availability.
Portals make use of the distinct logics of nonlinearity in their strategies and behaviors. Following from the focus of much of this final section on subscriber funding, the treatise now specifically explores this under-theorized revenue model and investigates the addition of a subscriber model to the flow, publishing, and written press models theorized by Miége because of its prevalence among portals in the marketplace.