The Flaws of "Subscription Fatigue", "SVOD Fatigue", and the "Streaming Wars" —


When we consider the state of tech-media in 2020, there are a few common narratives. The most inescapable is the “Streaming Wars”. In November, I argued that this term was a misnomer. Digital/streaming/OTT video is really just a battle in a much larger war: the “ecosystem war”. And for the most part, this war is fought asymmetrically. Apple and Amazon both sell digital media devices, third-party media content, and their own original content, for example. However, Apple isn’t an e-retailer nor a diversified enterprise cloud services provider, and Amazon doesn’t even have a smartphone, personal computer, watch or non-video app store. I’ll come back to this idea, but understanding the differences between these companies and their motivations is helpful when considering two other popular phases that are unhelpful at best and misleading at worst: “subscription fatigue” and “SVOD fatigue”.

“Subscription Fatigue”

The “subscription economy”, by definition, presumes that the overall “economy” – from products, to services, content, transportation, labor and more – is shifting over to “subscriptions”. Thus, to claim that consumers have “subscription fatigue” is to say that they have “spending fatigue”.

As always, most consumers will say they wish they spent less money, bought fewer things, and enjoyed lower prices. However, it makes little sense to say that the decision to buy TV subscriptions, radio subscriptions, toothbrush subscriptions, video gaming subscriptions, dog food subscriptions, car subscriptions, or productivity software subscriptions should drive “subscription fatigue” or mean each subscription competes with one another. For decades, consumers have bought TV, music, toothbrushes, video games, dog food, cars and Microsoft Office. What’s new is that they all have similar models – digitally-based, predominantly D2C subscriptions. This changes nothing about the individual value or baseline need for them.

Of course, the “subscription economy” does mean that step one of a recession will be to “re-evaluate all subscriptions”. However, this does not mean subscription fatigue should be considered a real “thing”, let alone a defining element of modern-day competition. Furthermore, payment model – upfront v. recurring, subscription v. á la carte, online v. offline – is irrelevant to what’s “re-evaluated” and not. Some subscriptions are “necessities”, like toilet paper, while others are concerned with discretionary spend, such as Office 365 or Netflix or Tinder. This latter group isn’t competitive because they’re “subscriptions”, but because there is, as always, finite spending money for non-essential items.

To this end, it’s important to highlight subscriptions are often a preferred buying path for consumers. Most would rather (or can only afford) $10 a month for a multi-year license to Microsoft Office for $300. Subscriptions also meaningfully reduce the cognitive burden of repeat decision making. No longer do you need to “track” your toothbrush for wear, risk “running out” of toilet paper and then be forced to overpay for a small-volume purchase, or need to scan and hoard coupons to ensure a great deal. Similarly, many consumers would rather marginally overpay for an all-you-can-eat subscription than optimize for specific tiers of use. In fact, most of us have caustic responses to per unit pricing, often to the point of irrationality (e.g. $40 for 35 loads of laundry detergent v. $1.00 per load). Amazon Prime is based on the need to get shipping fees out of the way once, versus fight them over and over and over and over, even if the effective shipping cost went up for a consumer, or the lack of shipping costs led to unnecessary purchases.

The rise of fully flexible monthly commitments also means that consumers no longer have to worry about having made a bad decision and being stuck with it. In this sense, every subscription is still á la carte, but unlike in the analog era, the default outcome of “doing nothing” is to keep getting value you enjoy rather than running out of a thing you need.

(Note that none of this means that a digital subscription business is a “good” one. Many sub-categories of CPGs and foodstuffs, not to mention music or fitness equipment, weren’t good business before the shift to subscription. The fact they shifted to subscriptions doesn’t inherently change this, just as it doesn’t mean they suddenly compete with all other subscriptions).

“SVOD Fatigue”

Of course, the nuances of “subscription fatigue” is separate from the question of “Subscription VOD fatigue”. It is obvious consumers don’t need 20 Netflixes. However, they’re not being asked to buy 20 Netflixes. It’s wrong to treat Fox Nation, Netflix, ESPN, and Twitch as competitors, let alone interchangeable “units of SVOD”. They serve very different functions and offer very different content. Just as Spotify and the New York Times and Amazon Prime shipping each do.

Amazon and Apple TV+, meanwhile, aren’t Netflixes – not in monetization, content volumes, or strategy. Now, if Amazon or Apple’s SVOD services can monetize so dramatically better through the Prime and iOS ecosystem than Netflix can via direct consumer spend and a singular focus, they can, in theory, “kill” Netflix – should they so choose – but that has nothing to do with SVOD fatigue nor the number of viable SVODs.

The question of SVOD fatigue isn’t about “how many SVODs will the average household have”. It’s really about “how many different roles are there to be played in video”. And the answer here is mostly path dependent – it depends on the innovation, risk taking, and discovery that happens in the marketplace, as well as timing. No one knew “live streaming video games” was an opportunity until Twitch, for example. And while Twitch likely steals time away from the video ecosystem, the viability of the Twitch subscription doesn’t mean that the number of viable OTT services has reduced.

The Question

All of which is to say what matters in SVOD is simple and not unique to SVOD: A service will succeed if (1) it addresses a real, outstanding customer want/need; (2) at an appropriate price or value to the consumer; and (3) while generating sustainable economics.

Quibi is a good example here. The company believes that there is an outstanding need for a new type of content, focused on a different time and place, under a different viewing behavior and focused on a specific audience. If it is right, and it can build up a defensible leadership position before other players replicate it, a new subscription will be possible and it doesn’t matter how many SVODs a customer already has (just as whether they have NYT or Spotify doesn’t matter). But of course, if you ask consumers “do you wish you had fewer subscriptions” or “fewer SVODs”, they will say yes – especially if they don’t really know what the new “thing” is. Note, too, that Pay-TV studies have been promising that 10%–30% of subscribers will cancel each year. They never do… because enough value remains.

More broadly, this three-point framework is well established (it actually has nothing to do with video). Over the past forty years, we have seen countless examples of “networks” launching into hyper-saturated marketplaces with hyper-specific but unproven (and often openly derided) theses regarding outstanding consumer wants and needs. Almost all of these have succeeded. In fact, they usually spawned several direct competitors – showing that the unmet want was even larger than original anticipated.

For example…

  • 1972: HBO launched a network focused on the most valuable TV time, Sunday night, with an unprecedented monetization model (á la carte consumer spend and no advertising), and focused only on reruns of Hollywood movies. It was ultimately bought by 25% of TV homes, became the most profitable network in the world and the market leader in quality. And this was despite the launches of Showtime (1976), Starz (1994) and Epix (2009).
  • 1977: Nickelodeon launched 24/7 content only for kids. No longer was kids content relegated just to afternoon and Sunday morning blocks. In the 2000s, Nickelodeon became the most watched cable network, despite having spun-off several other Nick-branded channels and seen the launch of The Disney Channel in 1983.
  • 1979: ESPN launched a 24/7 sports channel, ultimately with the highest programming budget in the world. In 2019, it brought in more than $2.5B in profits, with an annual revenue of roughly $9B. In 2009, Fox launched its own suite of 24/7 Fox-branded sports networks.
  • 1980: CNN launched a 24/7 news channel. Today, it generates an estimated $800MM a year in cash flow on $2B in revenue, and several other 24/7 networks exist.
  • 1981: MTV launched a 24/7 music video and culture channel that focused only on young audiences. The result was the first new Hollywood film/TV conglomerate in decades. Within years, MTV had launched several other 24/7 networks, while competitors launched even more focused versions, such as CMT.
  • 1983: BET launched a 24/7 network focused on black American audiences. In 2001, the company was sold to Viacom for $3B. Several other black-focused networks exist today.
  • 1996: Fox News launched a 24/7 news channel… only for half of news watchers. It now generates more than $1.5B in cash on $2.5B+ revenue

Of course, this sort of logic can be used to justify faulty assumptions around what opportunity exists, where, how large it might be, how durable it is, etc. In addition, these specifics gaps were open because of technological limitations. A network like ABC could only air one thing at a time – and therefore there were structural impediments to serving “everyone”. Netflix, meanwhile, can air anything, at any time, to every viewer, and on an individual basis.

But the crucial point here is that it’s wrong to think about the “number” of subscription video services, just as it was wrong to think about how “many” networks were in the cable bundle in 1980, 1985, 1990, and so on. In fact, it’s incredibly close minded to believe there are not more formats to be found. Or monetization models.

This last point is particularly important. Many emerging OTT subscription services hope to offer great value to the consumer through low/no pricing (✓ criteria #2) that nevertheless generates sustainable economics through their ability to drive other businesses (✓ criteria #3). This doesn’t directly solve the consumer need (criteria #1), but the good news about media is that consumers will always consider watching a new good show. This is especially true when these services are free to watch and part of an ecosystem a consumer already uses, like Prime.

But even when they’re not, consumers are open minded. After all, the TVOD market is still growing 20%–30% per year, which means consumers are still “happily” spending $5 on a movie rental or $20 to buy a new title. Starz is $9 and HBO Max $15, meaning a single SVOD month is basically a TVOD transaction with recurring revenue potential.

“Streaming Wars”

But to return to the monetization point, Amazon Prime Video is a good example. The service is given away free to consumers and most third-party estimates suggest Amazon is paying $5B+ per year on its content. This isn’t directly sustainable – but it is obviously indirectly viable in the “ecosystem wars”. To point, Bezos has routinely stated that Prime Video users buy more shoes and are more likely to buy and renew Prime.

More broadly, we can see clear and substantial evidence of the ‘Amazon Video’ flywheel. In In 2012, IDG reported iTunes had a 50% share of the $2.5B digital TVOD market, while Ovum estimated 70%. Today, some reports say Amazon has greater share globally, with the market having quadrupled in size. And while Fire TV launched seven years after Apple TV and eight after Roku, it is now the largest connected TV device globally. Reports suggest Amazon is now asking for 20%–40% of the ad inventory delivered over these devices. In 2018, TDG estimated that Amazon Channels, which sells third-party video subscriptions and aggregates them under its Prime Video app, was 53% of HBO, 72% of Showtime, and 70% of Starz OTT, and upwards of 75% of new subscribers. BMO estimates this will generate more than $3B in revenue in 2020, with 30% margins (far in excess of Amazon’s core business).

All of these drives clear revenue model even though it doesn’t require consumers to spend a dollar. Moreover, it’s not clear that because 150MM homes now receive a smattering of originals plus some library they’re now “fatigued” from adding other SVODs or other services have one fewer “SVOD slot” to fight over. In fact, Amazon’s success in aggregating has probably had the reverse effect! By making it easier for consumers to access, manage, and personalize their video experiences through a single interface, they’re likely to use and more support SVODs rather than tire after supporting two to four. The real fatigue from managing multiple SVODs is managing different authentications, bills and apps.

Apple, too, is offering its SVOD service for free or at low cost to consumers as part of the (immensely valuable) “ecosystem wars”. This effort is partly to rebuff Amazon, which per above, has stolen massive share in a category with clear economics. But to this end, Apple’s business case for spending $2B per year on original content is easy to see. But more broadly, it’s also clear that Apple is planning to make video a part of a much larger “Apple Prime” bundle that will span the company’s other content subscriptions (like Apple Music, Apple News, and Apple Arcade), software and services (Apple Care, iCloud), and hardware (the iPhone Upgrade Program). Again, content is part of the ecosystem wars.

The value here is easy to overlook but critical to understand what “finances” Apple’s investment in content (something many argue is wasteful, negligent, unsustainable, etc.). The Apple Prime bundle won’t just grow Apple’s service revenue and customer stickiness, it will also make the ecosystem “better” for its customers. Consider iCloud, for example. Many consumers still can’t come to terms with paying for cloud storage – but the iPhone and the overall iOS experience would be better if they did. However, because there’s a real marginal cost to cloud storage, Apple can’t just give it away for free. The same is true with Apple Care. Apple, its consumers, and iOS developers all lose if a consumer spends months with a cracked iPhone screen or broken home button, but again, Apple can’t do endless free iPhone repairs. However, if it bundles all of these offerings for free – many of which have no marginal cost and thus become cheaper on a per consumer basis as more subscribers are added – it can make its entire ecosystem stronger and generate greater overall revenue, even if the allocated ARPU per service (e.g. Apple Music) plummets through a bundle discount. This is more important, and sustainable, than scaling $5 subscriptions. To point: even if Apple amasses 100MM TV+ subscribers at $5 a month, it would see revenue and profits grow less than 2%.

AT&T has a similar model. The company didn’t spend $108B on Time Warner in 2018 to grow HBO revenues by $3B, cash flows by $1B and subscribers by 15MM by 2024. After all, AT&T already generates $182B in revenues, and $29B in cash with 166MM subscribers. In addition, this modest upside from HBO Max comes at the cost of +$5B in incremental investment and accelerated declines in DirecTV and Turner. In truth, a “telephone company” didn’t buy a “content company” to be in content. Instead, the acquisition is oriented around enhancing AT&T extant business units.

Every year, for example, AT&T spends billions to acquire customers (e.g. device discounts, promotions, marketing) and retain them (discounts). A 0.1 percentage point decrease in AT&T’s postpaid wireless churn (~1.9% in 2019) alone would generate an incremental $350MM+ cash profits. And so just as Amazon uses Prime Video to drive Prime adoption, engagement and retention, AT&T gives HBO Max away to high ARPU mobile customers and offers HBO Max discounts to lower value customers.

This sort of approach is uniquely possible in the digital era. HBO Max is a predominantly fixed cost business with nominal marginal costs. As such, it costs AT&T nothing to give $15 HBO Max away for free. However, it’s reasonable to assume the service has >$0 value to the customer. And every time HBO Max adds a customer, the per customer cost of HBO goes down (again, most of HBO Max’s costs are fixed), while the per customer benefit should be flat. It’s important to compare this dynamic to AT&T’s alternatives. Wireless discounts and device subsidies have direct marginal impacts on profits/costs/revenues. In addition, there are no economies of scale in customer discounts. Note, too, that when Verizon Wireless bundles Disney+ for free, it pays an estimated $3.5-4.5 per customer per month. There is a real marginal and non-scalable dynamic here, too.

Furthermore, AT&T hopes to use HBO Max to build up a large, multi-pronged digital advertising business (Xandr). Specifically, this will be powered by HBO Max’s data generation and ad inventory (in 2021, the service will add an ad-supported tier, which will likely be free to all lower-ARPU mobile customers).

The point here is that, as with Amazon and Apple, consumers are likely to get a “free” (#2) high quality (#1) SVOD service from AT&T specifically because of its (promised) ability to monetize indirectly (#3). And crucially, the wireless and digital advertising industries are enormously larger than that of subscription video. This isn't to say AT&T will generate a great return on $108B TimeWarner, but focusing on whether HBO goes from 35MM subscribers to 50MM subscriber, and whether that's "worth it" misses the point.

The final example that’s worth mentioning is Disney+. It has, obviously, been enormously successful out of the gate. And yet, it could be argued that Disney+ didn’t solve an outstanding consumer/want or need by yanking its content off Netflix. After all, tens of millions globally enjoyed watching films like Black Panther and Coco at no additional cost via a subscription they already had (Netflix). In fact, Disney might look like they created a problem so that they could solve it. However, whether this strategy “works” depends on far more than amassing tens of millions of subscribers. After all, Disney will have forgone tens of billions of dollars in pure profit, plus taken billions of additional investments in technology, content, and marketing, as well as considerable execution risk, in order to build this service. To point, the company doesn’t anticipate to breakeven until it has collected up to 90MM subscribers. That’s a huge hurdle.

For Disney to generate a return on Disney+, it must prove that repackaging all of its content into a single destination/brand/experience offers consumers a meaningfully “better” offering. There is a very plausible argument here for consumers: it’s a brand-safe environment for all things Disney, inclusive of unique navigational experiences, it includes integration into other Disney products/offerings, and is purpose built for super-fans and around IP ecosystems. In addition, Disney+ will offer something Disney could never get by selling its content to an intermediary like Netflix or Amazon: consumer data and a direct-to-consumer relationship. These two things should allow the company to make more intelligent programming investments, more effectively market its non-SVOD content and consumer products, and enhance upsell/cross-selling of everything from a Toy Story tote bag to a $5,000 cruise adventure. And crucially, this value won’t directly appear on the Disney+ P&L.

Although the “Streaming Wars” narrative dominated OTT video in 2019, the year was also defined by the ostensible rise of AVOD. January alone saw Viacom buy Pluto TV, Amazon launch IMDb TV, Hulu drop its ad-supported SVOD price from $9 to $6, and NBCUniversal announce plans for its own service which, in September, was unveiled as AVOD-centric, too. And in October, HBO Max announced plans to launch an ad-supported tier. ViacomCBS seems to be planning a similar offering that will converge Showtime, CBS All Access, Pluto and its other sub-brands, as well.

Some have argued such offerings are absurd – again, citing surveys. If no consumer “wants” more OTT services, certainly they don’t want their n+1 service to have ads! And yet a year later, Pluto, IMDb TV, Hulu with ads, and market leader Tubi all have more users and more watch time. The viability here stems from two aforementioned points – our desire to watch content we like and the monetization models that allow these services to be made available at little-to-no-cost.

Today, there is enormous pent up demand for rich, video advertising with plausible targeting. This is a result of the overall scarcity of ad-supported premium video viewing online (see Amazon, Netflix, Apple), the erosion of pay-TV overall (which has removed 25%+ of all ad impressions since 2010) and the fact that the dominant advertising medium, digital/mobile, is dominated by different ad formats (e.g. keywords). As a result, the value consumers place on ad-free experiences is (dramatically) less than the per consumer value generated by ad-supported experiences. We can see evidence of this at Hulu. The service charges $12 for ad-free, but generates roughly $15+ from its ad-supported tier ($6 in fees plus $9+ from advertising). This dynamic is so great the company gave up $500MM in annual revenue by dropping its ad-supported price in order to encourage ad-free users to “downgrade” and ensure new subscribers picked the ad-based tier. Note, too, that advertising revenues also scale directly with incremental viewership, while ad-free services benefit from more diffuse benefits (e.g. marginal reductions in churn rates and marginal increases in willingness to spend).

To summarize, the enormity of the advertising opportunity means that many consumers will receive subsidized and/or free “SVODs” thanks to the Fords and P&Gs of the world. This doesn’t mean tens of millions of households will love or use all of these services – you can still be “fatigued” of optionality and free things – but many consumers will. Especially when they’re easily accessed alongside through their aggregated UIs and devices of choice and full of good content.


It is the very fact that the streaming wars are just a battle mostly fought asymmetrically – and thus tap value differently – that so many can exist. To return to the subscription fatigue idea, your Prime subscription is not competing with your iPhone subscription – and as a result, it’s possible to get both Prime Video and Apple TV+ Originals without spending a dollar. On their video services, at least.

Certainly, this doesn’t mean there won’t be some losers, especially compared to the “old” Pay-TV system. It also doesn’t mean there won’t be increasing returns to scale, or that in a decade, the smaller players won’t be squeezed out. But that’s an economic and ecosystem “war”, and it’s different from “SVOD fatigue” and certainly from “subscription fatigue”. And asymmetrical warfare does not mean that these services aren’t still competing for finite attention. However, this only matters to the extent that incremental attention/engagement is what drives retention or the underlying business. Does Amazon need 30 hours per month to drive the Prime or Video flywheel, or is five sufficient? Does Disney need to beat Netflix’s engagement in order to charge a comparable price?

It’s also worth highlighting just how enormous the appetite for video is. Each day, some 300MM Americans watch an average of five plus hours of television. That’s a lot of “land” to fight for in a multi-party war.

Matthew Ball (@ballmatthew)

PS: Yeah from maybe 2013-2018 I perpetuated the war idea a lot. My bad.