A simple and powerful example of the shift from access to content-based competition is HBO. HBO launched in 1972 to provide ad-free access to Hollywood movies not long after they left theaters—hence its original title “Home Box Office”—at an à la carte fee beyond what regular cable packages cost. This was access-based innovation, but it was easy to replicate, as it ran on third-party television equipment (i.e., pay-TV providers) who were keen to sell more to their customers, customers HBO couldn’t sell to directly anyway. In 1973, another HBO-styled premium cable network, The Movie Channel, launched across the country, and a year later, Showtime as well. In the decades that followed, each of the major premium networks started launching spinoff networks, e.g., HBO 2, HBO Comedy, HBO Signature. This allowed each network to provide access to more content at the same time slot, thus reaching more customer segments. By the late 1990s, original content moved to the center of the model—HBO obviously noticing that bidding on externally produced box office films or PPV events was rapidly becoming a margin race to the bottom. HBO had begun its original programming efforts in the 1980s, licensing exclusive domestic rights to signature foreign series, such as the ITV series Philip Marlowe, Private Eye, starring Powers Booth, and the CBC series The Kids in the Hall (which moved to CBS after three seasons), as well as exclusive sporting rights, such as pay-per-view boxing. In 1995, HBO commissioned its first homegrown scripted series, Oz, which premiered in 1997. By the mid-2000s, half of HBO’s budget was being spent on original content, with The Sopranos winning Outstanding Drama at the 2004 Emmys. By 2013, 60% of viewers were watching these originals, and by 2016, 70% of on-demand viewing was HBO’s originals, not its licensed films. HBO was so successful establishing a brand and content expertise that by the early 2010s, the network was the third most profitable in the world, generating $1.2-1.5B in profits annually at a 40% cash flow margin.
HBO was a premium cable network and therefore emerged after broadcast television emerged and in tandem with basic cable, each of which tells its own story of access-to-content shifts.
When broadcast television emerged in the 1920s, it was a revelation for human leisure. It was the only at-home entertainment option that had both audio and video, making it easy to consume and more stimulating than the alternatives. Compared to print, TV was also “live,” thus making it more current and frequently refreshed. Video could also be consumed passively in the background, which made it convenient, like radio. But unlike radio, TV worked at least partly without sound (especially for sports) and also worked as an audio-only experience. Almost all broadcast content—at least during the primetime hours—could be enjoyed nationwide. As a result, broadcast television enabled, and counted, more monocultural moments than any entertainment alternative in history.
For nearly a century, television had a monopoly on spare time. The access it provided—multiple channels of live video inside the home!—was so profound it cannibalized time from basically all other forms of entertainment, including in-person socializing. And for decades, television’s leisure monopoly was held by just a handful of channels. The “access innovation” of broadcast could not be commodified because broadcast spectrum was finite, scarce, and government-distributed, thereby preventing new entrants. Furthermore, it was technologically impossible (or at least, financially impractical) to charge consumers for access or discriminate between different TV homes, thereby preventing the introduction/testing of other business models. The result is that the content-based wave of competition never really happened. That doesn’t mean content wasn’t important; it was why consumers watched one network over another at any given moment. But audiences could only pick from two or three practical alternatives on any given night, and naturally, these channels could not appeal to every prospective viewer at once. The broader and more popular a given show at 8 p.m. on ABC became, the more likely another show at the same time slot on CBS or NBC might appeal to those who didn’t love ABC’s hit. In other words, the ceiling for timeshare was around 50–60%, but the floor was 10–20%.
But throughout the back half of the 20th century, cable television networks began to spread across the United States. Though cable-based television content was still “live only,” it was suddenly possible to distribute dozens of channels to an audience accustomed to choosing among a half-dozen or so. The constraint was not technical but commercial.
But the growth in what could be distributed resulted in more being distributed and watched too, as audiences flocked to programming made for them, rather than the generic American or his or her entire family. And so MTV, which had launched in 1981 to focus on youth music culture, launched MTV 2 in 1996 to focus only on music videos; MTV Tres in 1998 to focus on Latinx Americans; and MTVU in 2004 for attending college. MTV’s parent company, Viacom, also expanded the kids-focused Nickelodeon, which launched in 1979, with Nick Jr/Noggin and TeenNick in 1999 and Nicktoons in 2002. Having successfully launched the Fox Broadcast Network in 1984, 20th Century Fox in June 1994 launched FX, a basic cable network focused on premium scripted series targeted at adults. In the months leading up to the release of FX, the company announced plans to launch FXM, which would distribute Fox’s adult-focused films through basic cable. And then in 2013, Fox split FX up into FX and FXX, with the latter focusing exclusively on comedy.
The advent of satellite, DOCSIS, and IPTV in the 1990s meant that it was possible for a household to receive hundreds of channels, including both the West Coast and East Coast feeds of major networks, as well as browse extensive VOD catalogues (or use their TiVos/DVRs/PVRs to record live TV for later consumption) timeshift. These delivery models also enabled sophisticated electronic programming guides that also brought some order to this ever-growing buffet through notifications, calendars, and consolidated search.
All of these technological advancements led to the so-called “Peak TV” era—the advent of content-based competition. Between 2002 and the premiere of House of Cards in 2013, the number of original scripted series surged from 82 to 336, or 310%. AMC, for example, followed the HBO model with great fidelity. The network had been built around reruns of old movies—hence the channel’s original name, “American Movie Classics,” not unlike HBO’s original name “Home Box Office”—but dove into best-in-class originals, ultimately making two of the best dramas in history: Breaking Bad and Mad Men (both of which premiered in 2007). Even Discovery Channel began producing high-end scripted epics.
Netflix, too, followed the access-to-content shift. For its first four years as a streamer, Netflix relied purely on its access innovations: Internet-based video available on demand, for a single all-you-can-eat price, with thousands of fully stacked seasons of TV, and thousands of additional catalogue Hollywood films. The company also grew through classic access-focused tactics such as ubiquitous distribution (the Netflix app was available on everything from top treadmills to the Nintendo DS), distribution-focused partnerships (buying placement on the Roku and TiVo remotes, wholesaling its subscription to T-Mobile), and well-below-cost pricing designed to drive adoption. But clearly understanding the coming phase in which its suppliers would take back their content and launch competing streaming services, Netflix in 2011 began investing in its own originals, which began to be released in 2013. By 2018, 85% of Netflix’s new spending was on original content.
The “Big Media” giants launched their streaming services five to six years after Netflix’s first original series and four to five years after Amazon’s. On the whole, these players believed that providing access to their catalogues and maybe a smattering of originals would be sufficient to succeed in the new streaming market. They would rapidly be disabused of this perspective.
Disney originally forecast that spending on Disney+, Hulu, and ESPN combined would be $2.5B per year ($1B service original + $1.5B licenses/catalogue content), growing to $4.5B ($2.5B + $2B) by 2024. In Q4 2020, only 14 months later, Disney nearly doubled this budget to $8–9B. Then in Q4 2021, Disney+ shocked investors by announcing that 2022 would see total company content investments grow by $8B year over year, from $25B to $33B.Paramount, meanwhile, said in 2021 that their 2024 direct-to-consumer spending would be $5 billion (not $4B as it had guided{?} in 2020), then a few months later upped the sum to $6B. In 2020, NBCUniversal said Peacock would have a content budget of $2B over its first two years. In 2022, NBCUniversal announced that it would spend more than $3B that year alone, and $5B in 2024. As I mentioned at the start of the essay, Peacock’s quarterly EBITDA losses are 49% higher than its total revenue—and growing tens of percentage points faster, too, with content expenses making up the vast majority of these costs. When Apple TV+ launched, its budget was estimated at $2B per year. By its third year, that sum had tripled.
There were many drivers for this escalation. At the core, several players quickly fell short of their subscriber growth targets and/or suffered from higher-than-anticipated churn. The proximate cause was that signature series simply didn’t connect with audiences as expected, a reflection of the intensity of content-based competition after 2020 (Peacock launched with NBCUniversal’s most expensive series ever, Brave New World, which was canceled three months later and has since been removed from the service). Similarly, many operators found that their catalogue/library delivered fewer subscribers than they had anticipated. While Friends and The Office consistently topped the streaming charts more than a decade after they finished their runs, there are even fewer of these than one might assume. Other episodic hit series from the 2000s and 2010s with enormous episode counts, such as Two and a Half Men, 30 Rock, CSI, Criminal Minds, and Seinfeld have not come close to these two NBC titans in the streaming context. And even Friends and The Office tend not to sustain long-running subscriptions—how many folks will pay $10–15 a month, every month, for access to the series (you can also buy the entirety of Friends and The Office on iTunes for $100 each).
Another cause of the escalation was basic game theory. There is a roughly 18- to 36-month delay from the time a service increases its programming budget and when the resulting films and series premiere on the service. As such, operators worried—and often saw evidence—that their competitors were ratcheting up their spending. The result was a feedback loop where more spending at Service A prompted services B–Z to spend more, thereby affirming each service’s choice to invest while also diluting the benefits from those investments, thereby requiring more investment.
SVOD, in other words, is a brutal business. In the pay-TV era, content-based competition was fueled by the fact a household had access to too many networks for a single price, which meant viewers could never consume everything they had available to them, could jump between networks at a second’s notice, and had no opportunity to save money by concentrating engagement. Furthermore, each network secured distribution to tens of millions of households for years at a time, thereby guaranteeing revenues and reach, which fueled content investment. But in streaming, the reverse dynamic prevails. Every network is sold standalone, month-to-month, and for as little as $5 and at most $15. And given the seemingly infinite abundance of high-quality programming, customers are incredibly promiscuous, inclined to join for high-profile new content and to stay as long as they think more is on the way.
But regardless of the specific causes, the result of content-based competition in pay TV, then Connected TV, then streaming TV is that there are now more than 800 original scripted series airing annually in the United States, up from roughly 85 two decades ago. What’s more, this figure excludes many foreign series, as there is no real “rule” for what is and isn’t counted. House of Cards clearly qualifies, but what about The Crown? The Crown makes sense, but why should it be included and not Germany’s Dark or Spain’s Casa De Papel? The intuitive answer is language of origin and greenlight rationale, but it’s strange to exclude Squid Game, which is Netflix’s most popular-ever series in the United States and was nominated for Outstanding Drama at the Emmys (it was not submitted to the International Emmys). But then how does one choose which foreign series to exclude? Choosing after the fact, based on popularity, doesn’t work. And so on.
Unscripted series have grown even faster, from fewer than 50 in 2002 to more than 1,100 today. The extraordinary growth rate stems from the low cost of production (an episode might cost a few hundred thousand versus $5–20MM for scripted series) and rapid timeline (a series can be conceived of, shot, and released in 6–12 months, versus 12–48 for signature scripted series). But the most obvious rationale for that growth is the sheer (and generally underestimated) popularity of unscripted content. Roughly 70% of primetime viewing in the United States is unscripted; for years, Dr. Pimple Popper beat Shonda Rhimes’s high-profile broadcast series such as How To Get Away With Murder.
The history of almost all media distribution models follows the access-to-content shift. In 2004, four years after its satellite-to-radio receiver service launched, Sirius made its record-breaking $500 million deal with Howard Stern in order to differentiate its offerings from those of XM, its competitor at the time—the two companies merged in 2008—as well as from terrestrial broadcast radio, satellite radio’s access innovations of sound quality, no advertising, and nationwide distribution having proved an insufficient lure for many commuters. When Apple launched its Spotify competitor in 2015, it launched with exclusive radio shows, such as The Zane Lowe Show, a clear contrast to Spotify, which offered only human-curated playlists. Apple also changed its free trial and payment policies to secure Top 10 artists, such as Taylor Swift, who had pulled her music from Spotify. Three years later, Spotify began buying exclusive rights to top podcasts (e.g., Call Her Daddy), podcasters (Barack Obama), and podcast companies (The Ringer, Gimlet), spending more than a billion dollars overall. It’s also no coincidence that gaming consolidation began in late 2020, after Amazon, Google, and Microsoft all launched global gaming services, while Facebook released the Meta Quest 2. New entrants in the delivery of content means exclusive content becomes essential. When Epic Games launched its own game store for PC in 2018, the company planned to differentiate by offering better economics and ownership rights to developers and consumers alike, as well as exclusive games. Not long after, Epic supplemented this strategy with a "free games" program whereby the company would let users redeem copies of paid games (e.g. Grand Theft Auto V, Death Stranding) at no cost. From 2019 through 2022, Epic gave away an average of 91 titles a year (worth an average of $2,000 combined in retail value), and users redeemed 600MM copies each year, with a notional value of $14B.
