Drivers of Television Consumption and Future Implications
Throughout my career, I’ve made or picked up on several observations on what drives television consumption beyond the relative appeal of specific content to groups of consumers.
Most specifically, I observe the following:
How content is accessed really matters.
Consumers won’t change how they access content or learn how to use the features their devices offer…
…Unless it’s really worth the effort.
Only a limited amount of content is “must watch”…
…And “Passive viewing” is much bigger than many people think.
Viewing share is roughly related to content spending share.
I elaborate on these ideas in the following note:
How content is accessed really matters. In the year 2000, I was a junior investment banker and worked a fair amount of time on industry landscapes related to the consequences of the then-recent Gemstar-TV Guide merger. Although few people likely remember the company today, it was potentially very consequential at the time. The basic idea was that in a world where cable and satellite operators were rushing to provide consumers with 500 channels of content and hyper-personalized experiences, the most important “real estate” in television would be the user interface. Gemstar provided interactive program guides to the industry and, with TV Guide content, owned the most important information that populated those guides. While Gemstar-TV Guide never lived up to those early expectations for many reasons (including fraud), I always thought their underlying premise was sound: how consumers search for, browsed or otherwise accessed content would be increasingly critical as content fragmented and as producers, studios or networks invested in new distribution channels.
Consumers won’t change how they access content or learn how to use the features their devices offer…A few years later, in 2003, we saw Disney launch Moviebeam, which was a dedicated set-top box intended to access Disney content. Using surplus broadcast frequencies licensed from local broadcasters, the company would pre-download content to be stored on a set-top box purchased by a consumer. Content, such as Disney movies, could then be accessed as rentals. Around the same time, an independent venture and broadcaster-funded company called USDTV launched with a similar technological concept, distributing a “skinny bundle” selection of broadcast and pay TV content. Both businesses effectively shut down by 2007, and while there were many reasons neither succeeded, one reason in particular stood out to me: the vast majority of US consumers already had pay TV at this time (penetration rates were approaching 90% of households), which typically included almost as many channels as John Malone foretold. However, those same consumers were apparently reluctant to learn how to use the “input” button on their TV remotes, and services which required using it had no traction. Whatever the benefits of new offerings, relative to alternatives, they were not enough to force a sufficiently meaningful change in consumer behavior.
…Unless it’s really worth the effort. By contrast, around this time in France, we saw the launch of digital terrestrial television, or TNT, and in the UK we saw the launch of Freeview. In those markets and elsewhere, newly licensed radio frequency spectrum was made available to deliver dozens of channels to consumers for free, excluding the installation of a dedicated set-top box. More commonly, consumers could easily acquire an inexpensive box that was an integrated TV receiver paired with a broadband modem from any of several telecommunications companies. In most markets where these services launched, pay TV penetration was low, and most homes only had access to a handful of channels. With such a massive difference in a TV experience, it was no wonder that consumers embraced a new approach to accessing television content. I still recall staying in a home in France around this time and being enthusiastically shown how to change the TV’s input by my not-so-technologically-focused host.
Only a limited amount of content is “must watch”…The DVR – the digital video recorder - was another technological innovation that was fast-emerging during this era. Much of the industry was literally freaking out about DVRs, but I recall that actual data from the UK’s Sky, where similar devices were launched much earlier, showed that only a relatively small percentage of all content available – usually the most popular prime time programming – was recorded by consumers who had accessed the machines for an extended period of time. And even then, commercial viewing remained high because often consumers couldn’t be bothered to skip the ads, despite the industry’s fears. This more or less played out in the US, suggesting that despite the fact consumers were watching an average of around four hours of content per day in the US, perhaps most of it didn’t matter all that much.
…And “Passive viewing” is much bigger than many people think. A few years later came another lesson for me: in what I understood to be the most extensive ethnographic research project around TV consumption ever undertaken, a Nielsen-funded but independently administered group called the Council for Research Excellence fielded several studies, including one from 2008 that looked at multi-tasking. As one of the studies revealed, only around 45% of viewing of live TV occurred when TV was the sole medium being consumed and also the sole activity undertaken. During around 35% of viewing, consumers in the study were concurrently engaged in a second activity, and during around 20% of viewing, TV either wasn’t the primary medium being consumed or wasn’t the primary activity a consumer was engaged in. Perhaps surprisingly (or perhaps not) the numbers weren’t radically different when we looked at consumption of TV during DVR playback (closer to 60% of viewing occurred with the playback as the primary activity). When we consider the actual quality of much of traditional television programming, it’s not hard to understand why: much of what airs on television networks is probably best considered as “ambient” or “background” content.
Viewing share is roughly related to content spending share. Overlapping many of these trends, one interesting data point I found many years ago was that if I looked at the amount of money that HBO spent on programming – a number regularly disclosed when it was part of Time Warner – and compared that number to the amounts spent by other programmers at the company level (most owners of networks didn’t disclose similar numbers at the network level), there was a decent relaItionship between HBO’s share of spending and its total share of viewing time. We can do something similar with data from Netflix if we compare it to viewing data tracked by Neilsen. Consider that last year Netflix spend around $17 billion globally last year in cash terms or $14 billion accrued on a GAAP basis. Let’s assume that the US allocated spending figure should be around $7 billion, and further assume that pay tv providers paid around $55 billion for their content. Cable and broadcast programming accounted for around 60% of all TV viewing during 2022 while Netflix accounted for approximately 7%. In other words, even considering for some fuzziness on the data and the question of whether or not one should focus on cash cost or accrued costs and focus at the distributor level or the network level, the relative amount of viewing per dollar of content investment will roughly hold.
Putting this all together, I note a few key conclusions when I try to anticipate how the industry will evolve in the US and around the world:
OEMs have a potentially big long-term opportunity when they serve as a primary consumer interface. When consumers “cut the cord” their primary point of access to TV becomes whatever interface the connected TV manufacturer enables. Any OEM who invests enough in this “front door” has the opportunity to produce significant value if they can help a consumer easily find content they want. Of course, merely directing consumers to programming solely based on sponsored results probably won’t serve an OEM as well in the mid to long-run.
Programming that is sufficiently good or important will cause an audience to find it. From the day that Netflix launched “House of Cards” in 2013, it became a worthwhile endeavor for an American consumer to figure out how to use their input button. As greater volumes of high-profile/ high-quality content was only available in this manner, it became even more worthwhile to invest time in understanding new ways to access this content For similar reasons, I think that if or when an emerging player chooses to make a meaningful content investment – let’s say Amazon buying NFL rights, which can only be viewed via the Prime app – the vast majority of consumers who want to see that content will find a way to watch it. Those who maybe didn’t want it so much – casual fans and “ambient” consumers of the content – probably won’t bother.
Lower viewing may be positive if it’s highly engaged viewing. Historically, the industry only focused on viewing as measured by Nielsen, and presumed it was all roughly comparable in terms of quality of viewership. By contrast, Nielsen data was never intended to identify whether TV viewing was primary, secondary or tertiary. I would argue that other than for live sports, in a world where DVRs became ubiquitous, DVR playback was going to be the best proxy for whether or not content was truly engaging and important to a consumer. To the extent that a growing share of viewing is consumed via SVOD or AVOD-based streaming platforms – almost as much as is consumed on cable and broadcast, according to data from Nielsen – it wouldn’t be surprising if total viewing eventually trends down, while quality viewing (i.e. the share of viewing without multitasking) goes up. Even at lower levels of absolute viewing volumes, higher quality consumption would be positive for the industry, as it likely reflect consumption that consumers are willing to pay for (and which advertisers should highly covet, when those consumers are available to be reached).
Content spending increases are only necessary if competitors do the same. Media companies don’t need to increase spending on content to grow necessarily, but if they fail to make such increases, they will then depend on others to cut if they want to avoid consumer time-share losses. If my observation above is correct, in a world where the WGA and SAG strikes are over but studios are more conservative in their spending, a studio/packager who cuts while others hold the line on budgets should expect lost audience share, and presumably revenue. Conversely, a studio/packager who holds the line or increases budgets while others are stable or declining should expect to grow on a relative basis. Putting it differently, we can anticipate with high certainty that linear broadcasting will never grow its consumption levels again, at least so long as studios continue to prioritize their content investments into content primarily distributed by streaming services.