Cloud gaming is the new “thing” in the media industry. Whether it’s a game publisher or platform, there’s constant talk and even more ambition to be the so-called “Netflix of gaming”. And notably, this opportunity is coming with a lot more momentum than OTT video ever did: Microsoft and Google will be launching their services by the end of 2019, Sony has already had its in market since 2014, Amazon is expected to join the fray in early 2020, publishers Ubisoft and Electronic Arts have also announced cloud-based subscriptions that will launch next year, with Activision Blizzard and Square Enix rumored to be planning the same. And so on.
This speed stems from several differences between the gaming and video industries. From a technology basis, penetration of high-speed Internet and personal media devices is much higher than in 2007 (when Hulu and Netflix’s OTT service launched), as is consumer willingness to spend on digitally delivered content. Furthermore, a meaningful portion of gaming spend and engagement has been online since the early 2000s and has been rapidly growing ever since. On the content front, engagement is also far more concentrated among a handful of titles in gaming than is the case in the video industry. This means that deep and broad catalogue rights are less important, and a plausible gaming service can be launched with only a few anchoring titles. And even where catalogue is concerned, these rights tend to be available. One of the reasons it took so long for Big Media to launch its streaming services was the fact that movie/TV libraries and future output tended to be sold for three to seven terms as many as four years in advance. And competitively, every media company has seen Netflix’s success since 2007, as well as consequences felt by its competitors who waited too long to enter digital video. To point, it has taken more than a decade after Netflix for the major content companies to launch their own D2C SVOD services, but many of the major game publishers are launching their own D2C services at the same time as aggregator platforms Microsoft/Amazon/Google.
Yet for all its enviable success, Netflix hasn’t really created new value in the video ecosystem; it just stole share from both content companies and distributors. In fact, it’s believed to have placed deflationary pressures onto the industry by cannibalizing consumption to an economic model that generates (far) less revenue per hour watched than traditional TV or transactional video, and accelerating cord cutting. Still, there are reasons to be more optimistic about the effect cloud delivery will have on the gaming industry. By the time streaming reached video, it was already the most popular and intensely consumed media category on earth. Gaming today has far lower penetration and less average usage per customers. Furthermore, gaming has always had far higher barriers to adoption and usage. There’s been no equivalent of free over-the-air broadcast in video gaming – gamers always needed to spend money, access a device, be in a specific place, and load up an experience. Cloud gaming will address much of this, and Twitch’s billions of minutes watched demonstrates how replicating elements of TV’s ease of use can drive substantial engagement gains. Google’s cloud streaming service, Stadia, hopes to use this type of consumption (but on YouTube) as an on-ramp to direct play.
Which is to say that the enthusiasm around cloud gaming isn’t just technical, nor about share gains or moving up/down the value chain – it’s also anchored around the belief it can drive a massive expansion in how many people play video games and how valuable the sector will become. Specifically, we hear lots about “the next two billion gamers” (Microsoft) or even the “next five billion” (Ubisoft). This latter figure, of course, is aligned with how many people frequently watch TV or have smartphones.
I’m not convinced that cloud gaming will deliver on this expectation – at least as traditionally defined and as applies to the gaming sector today. This doesn’t mean it won’t drive material incremental value – it will. It also offers a pathway to these customers, albeit not via traditional games.
But to get here – let alone to address to the two-to-five billion question – we need to start in the 1970s and with video game theory.
#1: Technology as Gaming’s Point of Origination
All media is the product/intersection of (1) technology (e.g. capturing, distributing, re-producing an image), (2) content, and (3) business models. Among other categories, video gaming has always had the greatest reliance on technology because ultimately a local processor has been responsible and required for the delivery of nearly every element of the consumer experience – from collecting local inputs, to rendering a live image, decision processing, outcomes analysis, and so on. Even in online gaming, controller inputs are gathered and at least partially processed locally, and algorithms are applied. Some of this occurs with music or video or text today, but computers/processing/reading typically became involved later in these mediums’ histories (e.g. CD players were the first to involve processors in music) and their roles were also lesser (e.g. content selection and presentation, rather than active generation).
This reliance has several interconnected consequences. First, gaming is more powerfully influenced by technology (and therefore technology change) than other categories. Put another way, business models and content flow more strongly from tech than to technology. Evidence of this is abundant; regardless of whether content or distribution is king, making games under the wrong technological premises will always come back to haunt you. Nintendo, for example, is widely believed to make the very best content in video gaming. However, this didn’t lead to the success of the Virtual Boy or the Wii U. In addition, the decision to use cartridges, not optical discs, with the Nintendo 64 allowed Sony’s first-ever console to surge into a leadership position and steal away key developers. Square Enix’s decision to leave Nintendo, its hardware maker of choice for more than decade, was based on its need for data storage capacity that only CDs could (then) provide. Conversely, Nintendo’s correct bets on technology, such as motion control with the Wii and portability with the GameBoy and Nintendo Switch, led to still-unprecedented success.
Second, the gaming category’s reliance on technology/computers means that it has experienced more frequent and more significant change than any other medium. This is particularly important because the technological change in cloud gaming is so massive. Among other benefits, it will be the first time in which local processing and rendering is no longer necessary and in which audiences can access supercomputers many times more powerful than they could ever afford to use, let alone buy.
Finally, gaming has been held back by the limitations of technology more than any other medium. But this same point has produced an interesting outcome: technology changes have always driven net additive growth. In other media categories, new generations of devices or formats have tended to replace the prior one (e.g. Blu-ray to DVD, DVD to VHS, Streaming Music to Digital Downloads to CDs to Cassettes to Vinyl, on-demand TV to cable to broadcast). In gaming, growth has piled up on top of each other like a geological stratum. This reiterates the idea that cloud gaming will grow the market.
To understand this specifically, we need to start with first principles. The “job to be done” of any entertainment medium is to pass time and provide leisure. Video games are no different.
But the specifics and context of this job matters. We want to pass time in different places, to different degrees, at different times and with different levels of engagement. TV is the best example of this point: its ability to flex to each of these different needs is a huge driver of its popularity. The average American watches nearly five and half hours of video a day – close to one third of their waking lives. This is only possible because of the possible range of this consumption. Some of it is purposeful viewing – we shut off the lights and ignore our cellphones while watching HBO Sunday night in the living room, for example. But most is distracted viewing – you’re cooking, on the treadmill, in a bar or working. This was further aided by video’s availability. Since its first release via terrestrial broadcast, video has always been anywhere and everywhere. It required nothing from the audience but a TV screen and click to access no matter where you were.
This latter dynamic wasn’t possible with gaming – a local processor was always required, content was always in a fixed location and set-up needed. Furthermore, games required full attention. You couldn’t just put it on casually in the background and there was no automatic ability to play elsewhere (the GameCube even had a carry handle, not that anyone used it). The video game sector was therefore able to keep growing through new formats and devices, rather than displace earlier ones, because each new device unlocked one or more of these limitations, while also offering different ways to interact with the category. We saw expansion starting from the very first major hardware shift: arcade to console
#2: Game Invaders
When arcades first emerged, we were still in the era of feature-specific computing. General purpose computers were too expensive and so most “computers” performed individual tasks: this is a word processor, this is a calculator, etc. Arcade consoles were the same – they played one game and one game only. And despite this singular focus, these devices were brutally expensive – often $2,000 or more (roughly $6,000 in 2019) – as they boasted (what was then) much of the most advanced computer hardware available.
Using top-of-the-line components might seem excessive, but among consumer technologies, video games have always tended to have the highest computation needs. At the advent of 3D gaming (e.g. Nintendo 64 and PlayStation 1), the CPU and GPU processing required exceeded that available in most high-end PCs, even those used commercially. In fact, the US government once feared PlayStation 2s would be linked together to enact cyber-attacks and orchestrate terrorism. The latency requirements between input (e.g. mouse or joystick) and image (screen) also far exceeded personal or standard professional use. And taking a step back, games required more diverse input and output signals to begin with, hence controllers.
As always, technology defined the audience, business model and content for a game. At $2,000 for only a single game, an arcade console’s only real hardware customer was a business (the arcade). This isolated the video gamer audiences to only those who so strongly wanted to game that they’d leave their house, travel to a shop and wait in line. The need to share devices also meant games had to be short and didn’t store progression data. Monetization, too, was pay-per-use (quarters).
The introduction of consumer-grade consoles in the 1980s therefore represented a ground-breaking change: suddenly you could game at home, play multiple titles, save (which meant much richer, longer, story-based games), and play endlessly without incremental payment per use. But this hardware was still expensive (The Atari 2600 was $800 in 2019 prices; the NES was $450) and could only be used to play games.
The launch of the PS2 and Xbox in 2000 and 2001 represented the transition away from game-dedicated devices. Suddenly, a gaming console also played video media (DVDs). Not long after, it was possible to play most console-grade games on general purpose PCs and laptops, whereas PC-gamers had previously required dedicated gaming PCs. And by the time the PS3 and Xbox 360 were mid-way through their sales cycle, they became the most common way to access Netflix on the living room TV (in 2011, the two consoles made up 55% of total Netflix usage according to Sandvine). The Xbox One, released in 2013, leaned even further into general purpose capabilities. It was designed to be an “input one” in the living room – the starting point for all families to access music, video (including Pay-TV, which would route from a provider set-top box into the Xbox One) and audio in the home.
A few years later, mobile computing devices began to grow as gaming platforms, setting the stage for computer games to be driven by truly multi-purpose devices that were accessible to all and, unlike the Xbox One, not even built for gaming. The most successful games in this category were free-to-play, meaning it was no longer necessary for players to invest in or even spend on either hardware or software. This led to a quick and massive expansion in industry revenues and consumers.
#3: Growing Up into the Cloud
The forty-year history of the video game market is one of three constant trends. And at the core is technology, which changed business models, content and consumers. At the dawn of arcade games, video gamers were mostly serious video game lovers – those who would go out of their way and spend a la carte to game in a public place. In the console era, gamers invested dedicated game machines for their private, at-home use. This decision was about how to spend leisure time – whether to game versus watch TV – and this decision came at a material upfront cost; it couldn’t just be made on a Sunday afternoon.
It took 25 years until the industry hit casual gamers. And this last wave of audience and revenue growth may look similar to those that preceded it, but it’s crucially different. These new players played not because they were gamers, or even liked gaming per se, but in part to pass idle time or because they couldn’t do other things they wanted to do (e.g. make a phone call in the subway). Indeed, part of this player expansion wasn’t even from attracting casual gamers, but social ones – those who game because others are, just like some of us who only drink or smoke because we’re with those who are. Early Western leaders in the category, such as Zynga, built their early userbases exclusively through the use of these social triggers (as facilitated by the social platforms).
This is why cloud is so important. New technologies have continually led to a dramatic increase in the gaming userbase. Cloud gaming will further expand access, remove friction, augment processing power and reduce cost barriers – finally making it possible for even the most casual gamers to access the best the category has to offer. Accordingly, it feels right to say it will lead to a substantial expansion in players and usage, too. But this is unlikely. To show this, some case studies are helpful.
A – Fortnite
Today, Fortnite is available in every region except China, at no cost, and on any computing device that someone could conceivably play a AAA game on. In the developed world, this means PCs, consoles, and mobiles devices. In the developing world, this is mobile devices. From this perspective, Fortnite has been able to maximize its reach the fullest extent that is legally and computationally possible. Were the game to be purely streamed (v. installed locally and then played online), players would benefit from automatic patches (which would drive increases in the time spent playing) and graphical improvements – albeit in exchange for greater latency – but no one new would be able to access the game. It’s already free and ubiquitous.
If a AAA game such as Fortnite, which is already considered to be cutting edge, can run across almost all devices, have global reach, be fully free-to-play and already cloud-based, one has to ask how cloud gaming will deliver 2-5B new gamers. The answer may be that Fortnite is the issue (though it has a fairly simple 20-minute play dynamic); audiences want other AAA games such as The Legend of Zelda or God of War or Halo to be cloud-based.
However, it’s unlikely the mobile form factor is a good fit here. There’s a reason mobile games look the way they look, play the way they play and that the most successful titles tend to be the simplest (e.g. Candy Crush and Disney Tsum Tsum) or simplified versions of console grade titles (e.g. Arena of Valor or Clash Royale). These games are neither as sophisticated, nor as complex as they can be – and by design. They’re instead optimized to the device and to the audience. And to point, Fortnite no longer matches console/PC players with mobile (or even Nintendo Switch) players.
B – Nintendo Wii
As a case study, the Wii provides important context to the one strange, thesis-conflicting aspect of the earlier chart of industry revenues – the contraction of console revenues. Under its “game population expansion” strategy of the 2000s, Nintendo launched Nintendo DS and Nintendo Wii, devices that added a stylus and motion controls designed to make it easier to play games. In addition, they released many ungamelike titles, such as Wii Fit and Brain Age. This led to huge hardware and software sales, with the number of people who played and bought video games increasing substantially.
Yet five years later, you can see in the chart that the console segment reverted back to its original curve; the Nintendo effect chipped away. The company admitted this in 2012 when it incurred its first-ever annual loss, with President Satoru Iwata stating, “The Wii was able to reach a large number of new consumers who had never played games before by bringing hands-on experiences with its Wii Sports and Wii Fit. However, we could not adequately create the situation that such new consumers played games frequently or for long, consistent periods. As a result we could not sustain a good level of profit,” In fact, this move was value destructive. Iwata: “Moreover, regrettably, what we prioritized in order to reach out to the new audience was a bit too far from what we prioritized for those who play games as their hobby. Consequently, we presume some people felt that the Wii was not a game system for them or they were not willing to play with the Wii even though some compelling games had been released”.
Much of this decline stemmed from the rise of mobile. Mobile games were easier to use, easier to access and more innovative business model. For casual gamers, or even mid-level gamers, it was a better device and more fitting content for their needs – which is why it grew overall revenues so much larger and so sustainably.
This tells an important lesson. Access to AAA, hardware-based games isn’t the core issue. Millions of households (and thus even more millions more people) already had access to a richly powered console – they didn’t need to pay more to use it. But they nevertheless shifted elsewhere. They didn’t want AAA, immersive, skill rich games. They wanted easy-to-consume alternatives. In a sense, this is similar to the incorrect but persistent belief that audiences just want more The Crown and Game of Thrones and Mad Men. Most want Duck Dynasty and Keeping up with the Kardashians. There’s a reason mom doesn’t play Call of Duty on Xbox when Billy is at school but does play Candy Crush. Hardware access isn’t her problem.
C – Hardware & Subscription Economics
If the Wii case study is ignored or too flawed (maybe they just didn’t make enough casual games or the right casual games, or portability would have solved the problem, or it was bad timing given the launch of iPhone a year later), it’s worth considering the independent cost argument. Perhaps there is substantial unmet demand for AAA gaming, but the hardware cost is indeed too prohibitive.
However, cloud streaming isn’t cheap. Most of these services are charging $10-15 per month (and you may still need to pay for some games) – which is effectively more about financing/renting hardware than replacing an upfront investment in a $250 console (which ends up being cheaper after two years anyway). Furthermore, the data bandwidth needs to support high quality AAA-gaming experiences are intensive – 5-16GB per hour with a range of 10-15 Mbps. Most homes would need to raise their mobile and home broadband caps, as well as their speeds. That’s expensive. Cloud gaming, unlike prior technological advancements, isn’t taking cost out. And for most players, you’ll still need to buy controllers ($60+) and other hardware (Google Stadia requires the $70 Chromecast Ultra). (For what it’s worth, hardcore gamers might not mind the added cost for improved functionality – but cloud streaming increases play latency and endangers fidelity, two elements of gameplay that this segment of users tends to value most).
Furthermore, console hardware has never been cheaper than it is today. Adjusted for inflation, prices have fallen between 30-60% since the 1990s. An Xbox is only $225 today and comes with multiple bundled games, with many other market leading titles (such as Fortnite and PUBG) available in full for free. Compared to $1,000 iPhones and cloud streaming subscriptions, console hardware is inexpensive. It’s also sensible even if you want to use cloud streaming services. Not only will owning a PlayStation or Xbox improve the cloud streaming experience (and obviate the need for separate controller or streaming hardware), Sony and Microsoft do/will allow console owners to cloud stream games they own to other devices for no monthly fee. What’s more, 50% of American households already have a major console. It’s hard to imagine millions want a console, or to play Tomb Raider, but haven’t because of the hardware fee.
#4: Cloud Gaming Will Still Grow the Market
The above case studies don’t mean that there isn’t value or upside to reducing friction of playing or accessing games. There is. Imagine needing to spend five minutes waiting for your television to load, or waiting fifteen minutes to update to this week’s HBO, or being unable to watch TV on your daughter’s television or at your friend’s house, or running out of “space” on a TV to add Showtime. That’s video gaming today.
Cloud gaming will make it easier to start playing and in more places – all the while maximizing actual play time. This will grow both play time and revenues. Fortnite’s success is fundamentally dependent upon the frequency and significance of its patches and updates. But as a free-to-play game, there is a real cost to the fact that for at least fifteen minutes a week, players can’t play and can’t transact – they can only wait as an update downloads and installs. That said, there are also technical solutions that aid much of this problem, such as automatic downloads, or downloading only the minimum required to start playing, with the rest happening in the background.
Subscription-based access, which makes incremental games free, and the ability to play a new game right away (versus going to the store/ordering it from Amazon/downloading it for forty minutes), will also drive gamers to sample and thus play more titles. This is healthy for any industry, especially one with such concentrated usage and sales. The addition of bundled subscriptions overall will also enable many more games to survive – namely those that were a poor fit for free-to-play monetization but couldn’t have survived at $30-60 prices. Finally, publishers and platforms alike will benefit from the ability to take back value from the grey market of game reselling from Craigslist and GameStop, as well as software borrowing. The total value here isn’t immense (GameStop alone does $2B a year in resales, with estimates for Craigslist and other sellers adding another $2B+), but it will dramatically grow bottom-line profits.
Each of these benefits is real and significant. But they’re not arguments for unlocking new groups of players; they’re arguments for better monetizing the existing audience of AAA-playing, hardware-based gamers. And secondarily, it’s offering publishers the opportunity to disintermediate (or take some share back from) platforms and to improve their customer data.
#5: How Could Cloud Gaming Reach Billions
All of this returns to the thesis that mobile is already maximizing the gamer audience, and mobile products are largely the result of what the market has demanded rather than what processing capabilities and input capabilities are or aren’t available. The complexity of game design and the skill and attention requirements will always limit the AAA gamer audience and their playtime. Cloud gaming solves neither of these issues – nor is the added computational horsepower offered by cloud delivery particularly relevant for casual mobile games.
Most of those in their 30s-50s who don’t play console/PC games and have never been interested before will not suddenly become gamers because they no longer need $300 hardware (again, millions of these customers already have them in their living rooms, too). While the console cost barrier is prohibitive in the developing world, cloud delivery doesn’t solve for this – access fees and bandwidth is still expensive. Even Netflix, which requires no incremental hardware and uses far less data than cloud gaming, has reached only 10MM subscribers in India, or 4% of households.
Cloud gaming will be successful. It will grow the market and it will reach tens of millions. And during this time, the industry will also see substantial growth in the number of gamers. But this will mostly stem from long running trends of gaming penetration. Cloud technologies, unlike prior format changes, have yet to meaningfully reduce the consumer costs or demonstrated new types of gaming content. In fact, we haven’t even seen a presentation on cloud gaming really discuss what new content experiences are possible. And the most cloud-based games today, such as Fortnite, are largely identical over such technology.
This is a shame. The benefit of cloud gaming, at the end of the day, is going to be in the types of content that could only be created via the cloud – the same way there was value to the types of games that could only be created on console versus arcade, or touch screen versus console. And this has always the case in media.
#6: Non-Player Participants
To start, one needs to anchor on three ideas: (A) What makes sports popular; (B) What makes TV popular; and (C) What cloud gaming is uniquely possible of doing.
A – What Makes Sports Popular
The most popular sports globally tend to be the simplest to understand and to play: soccer, cricket, chess, football. Being fun to watch and well-designed matters – but ease of access (which allows people to play) and intuitive rules (which allows non-players to enjoy it) are particularly important. Unfortunately, console-based video gaming is always much harder to follow, more complex in construction, and require more minimum skill to enjoy than traditional sports. This, again, is why casual gamers play casual games even when they have access to consoles.
This connects to sports viewership. People who enjoy sports on TV do not actually have to play that sport. Today, essentially all esports viewers play the games they watch, and those that don’t play also don’t watch. At its core, this is because these games are made for players – not viewers. Imagine watching baseball, for example, only from the helmet camera of the batter. That’s essentially how esports works today. To this end, it’s no coincidence that esports sprouted from fighting games. Unlike other genres, fighting games engage a spectator camera view and include on-screen graphics (such as life bars) that make it easy for any viewer to see who is winning. This also helps explain Fortnite’s domination on video game broadcasting platforms – the goal of the game is simple, the only information needed is plainly displayed on screen, and the camera set in second, not first person. To point, sports broadcasters have spent decades doing everything possible to improve the viewer experience, from digital first down lines in American football, to digital strike zones in baseball, to ill-fated tests highlighting a hockey puck.
B – What Makes TV Popular
And TV, as has been discussed, is successful for many reasons, but central to this is the ability of a viewer to modulate their levels of attention investment. They can lean-in, lean-back, lean-out, work, clean, workout, immerse themselves, and so on. Gaming doesn’t quite have this flexibility. Casual mobile games offer it to some degree – you aren’t really immersed in Candy Crush, nor is the cognitive burden overwhelming, but consistently attention is still required. And certainly, AAA games aren’t designed around engagement flexibility in the first place. Instead, they persistently require substantial investments in skill development, time and immersion.
C – What Cloud Gaming Can Uniquely Do
What’s most interesting about cloud game delivery is less about the ability to deliver AAA-powered games anywhere and without local console hardware, but how it can address the above issues. It’s not presently possible for mass groups of players to operate concurrently in the same digital environment (Fortnite’s Marshmello concert was really 100,000+ instances of up to 100 players participating in simultaneous reproductions). But it is possible to have millions of spectators watching concurrently and providing limited input interactivity into a live game or event.
The next step for gaming – and for esports and even live real world games/events – is to create content that is architected around audience input, influence and consequence. Here, the audiences does get to “play”, but not fully, not immersively and not with their full attention required at all times. This type of experience will be similar to the titular Hunger Games – there are contestants, but it’s not really about them; rather, it’s about the audience. And the audience can send in items, influence the stage, defend its favorites, and undermine those it dislikes. This could come from as little as a tap – a vote, an item selection, a thumbs-up – or as much as a casual/mobile game-styled minigame that would intersect into the main event.
This may seem radical, but this overall trend has been in place for more than 20 decades. At first, game design was a work of predetermined creation – the developer created every aspect of the game and users simply relived the world according to how the developer organized it. But since the first user-generated maps and mods, design has been transitioning to a model where the developer merely determines the guideline/code and establishes a system – similar to nature – that enables autonomous changes and for users to find their own way of playing. The transition to mass, spectator-influence is the next step. And notably, Fortnite-maker Epic Games is already investing here. Leaks suggest that they will soon release standalone mobile apps that allow viewers to watch and talk to their friends while they play.
And yet, it will assuredly be resisted for years to come. Ever since consoles emerged, the most recent generation of game developers have expressed skepticism about the new platform and sought to justify the superiority of the current one. Arcade game developers, for example, looked at consoles and saw weaker processing capabilities and generalized controller inputs (e.g. racing games, piloting and fighting games all had the same controller – not a wheel v. yoke v. button array). This, of course, led these companies to simply port their titles to the consoles, which allowed new developers with a console-first mindset to become the new market winners. Nintendo, for example, saw not constraints in consoles, but a newfound ability to tell lengthy, narrative stories – and they produced games that matched this opportunity with standardized controllers (Mario, for example, was a side-scrolling platform game that only required a D-pad and jump button).
We’ve seen this dynamic play out countless times. Companies that led in offline consoles, for example, aren’t the same companies that led in online gaming, and aren’t the same companies that thrive in mobile free-to-play. Fortunately, the gaming industry has never been one that destroys developers or platforms just because they didn’t lean into the next gaming category.
The other large opportunity in cloud gaming extends upon the idea of spectator-participants but feels even less like “gaming”. Instead of a “game”, it’s the creation of (buzzword alert) real-time, interactive and persistent digital worlds that offer players (or “users”) a sense of individual personhood and physical presence. For this type of experience to work, it needs to be accessible from any place, on any device, with any input and without latency – and with much of this experience managed and operated live in the cloud. This sounds like a somewhat fantastical abstraction as we don’t really know what this will or won’t do, why it’ll be fun, how it’ll work or even what it might look like. But it’s likely some ultra-creative studio with no real gaming history will find a concept that jump-starts this “category” with an idea that feels new and different – just like Niantic did with Pokémon Go, and Mojang with Minecraft. In fact, this is already where Fortnite, or more accurately, Epic Games is heading.