Satya Nadella has grown used to the naysayers. For years, Wall Street analysts questioned why Microsoft, the company famous for its Windows operating system and Office business suite, would waste money on something so seemingly trivial as video games. The calls grew louder when Nadella took the company’s helm in July 2014. Still smarting from his predecessor’s missteps in mobile devices, Nadella promised to steer Microsoft away from consumer distractions and toward its highly lucrative business services. Some even urged Microsoft to exit the gaming business altogether. “Four to five years ago, we and others were calling for them to divest that piece of the business,” says Daniel Ives, managing director of Wedbush Securities and a longtime Microsoft observer. That tune has changed: Last year, Microsoft’s gaming revenue—which includes Xbox, Windows games, and a cut of third-party gaming sales—topped $10 billion for the first time.
When I ask Nadella why the company didn’t drop gaming, he chuckles. “There were a lot of things that a lot of people said Microsoft should be doing,” he says. “If I listened to everything that everybody else on the outside asks me to do, there would be very little innovation in this company.”
To be fair, in years past, Nadella had been hesitant to call gaming business core to Microsoft’s overall strategy. Despite its success, gaming represents about a tenth of Microsoft’s annual revenue. Cloud-computing growth is a big reason that the company’s market capitalization topped $1 trillion this year; its “intelligent cloud” unit, which includes its Azure cloud-computing service, generates as much revenue in a quarter as the gaming group generates in a year. (Hasta la vista, Halo!)
But what if you could hitch gaming’s fortunes to Microsoft’s potent cloud engine? Well, now you’re talking. Nadella’s blockbuster $2.5 billion acquisition of the enormously popular world-building game Minecraft in 2014 was a “bit of a head-scratcher” when it was first announced, says analyst Ives, but it’s now clear that the CEO was “planting the seed of how he viewed gaming as part of the broader business.” Microsoft wouldn’t just retain video games. Much as the company managed with Windows and Office, it would use the flywheel of its cloud-computing infrastructure to dramatically boost the scale of its gaming business—and the fortunes of every video game publisher it works with—far beyond what was previously possible.
Today, gaming is unquestionably “core”; in late 2017, Nadella elevated gaming lead Phil Spencer to the company’s executive leadership team to underscore the point. And executives are bullish on the prospects of cloud-driven gameplay. Julia White, who leads product management for Microsoft’s cloud platform, estimates that the business of selling Azure services to video game publishers is worth $70 billion—about as much as publicly traded transportation darling Uber. Most of today’s Internet-connected video games are developed in, and operated from, private data centers run by game publishers, she says. Technology trends in other industries suggest that won’t last. “Even though game developers are in a very different business,” she says, “they face the same trials and tribulations of a commercial bank or a retail company going to the cloud.”
To the cloudmaster go the spoils: In January, the Xbox maker shocked the gaming world by landing longtime console adversary Sony (of PlayStation fame) as an Azure customer with a promise to collaborate on future unspecified gaming projects. It was as if General Motors and Ford had announced a partnership to take on Tesla—an unmistakable sign that the competitive landscape would rapidly and dramatically change.
It was also an indication that Nadella’s mission for Microsoft would be more expansive than it originally appeared. When I ask him why Microsoft is working so hard to build a consumer entertainment service when it has positioned itself as an enterprise software company, he replies, “It’s a bigger business, right? It’s bigger than any other segment. Why would I not do gaming? It fits with what we do. It has connective tissue to the common platform. We have a point of view that what we can do is unique.”
The problem: so does every other player in this game.
For 39,000 viewers tuned into Twitch, Elvis might as well have entered the building. Richard Tyler Blevins, the 28-year-old celebrity “streamer” known to fans by his moniker Ninja, has logged on to the service to play a few public rounds of the popular “battle royale” game Fortnite with his buddy. As his avatar runs and leaps through the game’s virtual environment, weapon in hand, Blevins barks commands like an NFL quarterback at the snap—and his Twitch viewers hang on every mundanity. Their comments rush by in the chat window accompanying Ninja’s feed. Some viewers respond to every move Blevins’s character makes (“get that delay ninja”); others practically ignore the show to talk among themselves. (One thread of conversation among many: Why Finding Nemo was a “pretty good” Pixar movie.)
In other words, just another day on Twitch. Viewers—overwhelmingly male and mostly 34 or younger—watched a breathtaking 9.36 billion hours of gameplay on the platform last year, according to estimates by production company StreamElements. Twitch launched in 2011 as a spinoff of streaming video site Justin.tv, a pioneer in user-generated content. In 2014, Amazon reportedly spent $970 million to acquire the site, besting YouTube-owner Google in a bidding war. Wedbush analyst Michael Pachter estimates that Twitch brought in $400 million in revenue last year.
Twitch, which is housed in Amazon Web Services, the online retailer’s cloud-computing unit, has rapidly become a cornerstone of the company’s broader video gaming strategy. AWS, as Amazon Web Services is known, is already selling computing resources and developer tools to video game publishers. It’s also rumored to be working on a service that would allow it to stream video games themselves rather than merely video of people playing them. (The company declined to comment, though recent job listings for technical roles for “an unannounced AAA games business” suggest its intentions. Like minor league baseball, “AAA” denotes the highest level of play in terms of budget and production.)
Two major milestones in the gaming industry set the stage for a cloudy future. The first: The massive success of Epic Games’ Fortnite, which brought in an estimated $2.4 billion in sales last year and now claims 250 million registered players. Fortnite demonstrated that “cross-platform” games, playable across competing devices from Microsoft, Sony, Apple, and others, could amass audiences far larger than those of the previous era, when titles were limited to specific ecosystems. “Fortnite was critical in getting the message across to all platforms that they have to lower the barrier of entry to their respective walled gardens,” says Joost van Dreunen, head of games for market researcher SuperData.
The second? Twitch. The service demonstrated that people were just as happy to watch and cheer people playing games—call it the kid-sibling phenomenon—as they were to play the games themselves. That kind of interactivity proved that engagement and gameplay were not one and the same. The dynamic expands the addressable viewership for a given title. “Viewing is eclipsing gaming, and a lot of youth of today would say they played the game when they really viewed the game,” says Bonnie Ross, head of 343 Industries, the Microsoft studio that develops Halo.
For Microsoft’s part, the company never saw the spectatorship aspect coming. “Amazon has Microsoft on a treadmill,” a former executive says. Two years after Amazon bought Twitch, Microsoft acquired competing service Beam for an undisclosed amount. Rechristened Mixer, it has become the means by which Xbox customers can watch one another play games, logging 39.6 million hours of viewing in 2018, per StreamElements—a whopping 179% more than the previous year but still a distant third to Amazon’s Twitch and Google’s YouTube Live.
The summer sun blazes above the thousands of coders assembled for Google’s annual I/O developer conference in Mountain View, Calif., but the anxiety on display in the long line has little to do with the weather. The event’s attendees, who base their livelihoods on building software for as many users as possible, are keen to hear Google’s sales pitch for why they should create games for Stadia, an experimental cloud-gaming service that the search giant promises to debut in November.
Like most Silicon Valley presentations, the executives onstage overwhelm with ambitious assurances of technical prowess. Stadia’s complex cloud architecture will prevent the nasty networking hiccups that cause online gamers to throw down their controllers in frustration, Google’s representatives say. All gamers will need to do is open a tab in the Chrome web browser; with just a few clicks, they can play a high-speed, high-resolution title such as Assassin’s Creed Odyssey.
Like their counterparts at Microsoft and Amazon, Google brass believe their vast data center empire gives them an edge on the technical demands of streaming high-end video game titles without interruption. Like its peers, Google has encouraged its consumer gaming and enterprise cloud groups to work together to ensure Stadia launches without the problems that have traditionally plagued online games.
Thomas Kurian, a longtime Oracle executive who is now chief executive of Google’s cloud business, says the company’s enterprise engineers built the networking technology that powers Stadia. Cloud gaming is a way for Google to penetrate a multibillion-dollar industry, Kurian says. “Our hope is that it’s expanding the market, not just being a replacement market,” he says. “For every person in the world that games on a professional desktop, there are probably three who can’t afford one.”
In other words: Why fight over a quarter of the market when the rest is greenfield? John Justice, a Microsoft veteran who now leads product development for Google Stadia, agrees. Gamers no longer want to “buy an expensive box every few years,” he says. Stadia, and services like it, are more accessible destinations to engage with games without the high barriers of entry found in the traditional console market.
Even the pricing plays a part: Though Stadia’s $129 bundle plus $9.99 monthly subscription has already been announced, Google says it is also evaluating a free version, with lower-quality graphics, that would debut later. Though the technological trajectory is clear, it’s still “early days” for the business model behind cloud gaming, Justice says. “Some people really do want transaction models, and some people want subscription models,” he says. “I don’t think we will say we will only go with one.”
It could take years to iron out the details. Though consumers would love a gaming model akin to Netflix or Spotify—pay a monthly fee, play titles to your heart’s content—it’s not yet clear that cloud providers have the leverage over game publishers to make that happen. Publishers have seen how platform pressures have changed the business of movies, music, magazines, and more. They don’t want to give up a share of their sales unless they’re certain that there are many more to be had in the long run.
Ubisoft, the French publisher best known for the Assassin’s Creed series, isn’t terribly concerned. “That’s less interesting to us,” says Chris Early, an Ubisoft executive who manages partnerships and revenue. The company in June revealed its own subscription service, called Uplay+, that is playable on personal computers and spans more than 100 titles in its own catalog, including Far Cry and Prince of Persia. It costs $14.99 a month and will also be available on Stadia next year. At this moment, “it makes less sense for a publisher to be part of an aggregated subscription model,” says Early. There are many proposals for how to sustainably monetize cloud gaming, he adds, but it remains unclear “who is going to pay whom.”
For now, publishers are focused on figuring out whether today’s successful titles make sense in the cloud—or whether all-new titles, native to the format, will replace familiar franchises. The interactivity of Twitch and the novelty of so-called freemium mobile games, like Candy Crush, showed that technological leaps could open new paths to gaming engagement. The possibilities that could emerge from running games on the same infrastructure that supports today’s artificial intelligence are something that technologists can only fathom.
“There will probably be evolutions of game design that we can’t even imagine yet,” says Early, “and they’re going to take advantage of the increase of cloud compute.”
Back in Redmond, I stop by Microsoft’s 343 Industries game studio, where employees welcome me to a visitor center—a shrine, really—celebrating the company’s Halo franchise, which has racked up $6 billion in sales since its debut. Statues depicting its heroes and villains tower over my head—a gallery of Greek gods, so to speak, for the gaming set. There are glass museum cases everywhere packed with memorabilia. On one wall is a rack of replicas of the virtual weaponry from the game, as intimidating in person as they appear on the screen. Bright orange tags with the word “prop” hang from their triggers in case someone takes the “incineration cannon” a little too seriously.
Founded in 2007 and named after a Halo character, 343 Industries is one of the older members of the Microsoft game portfolio. Last year alone, Microsoft acquired six game studios; at this year’s E3 industry confab, the company announced that it had picked up one more. Today, its Xbox Game Studios division is a federation of 15 semiautonomous studios that the company believes will be a key asset in the cloud-gaming wars—particularly against Amazon and Google, which lack strong titles of their own.
Not everyone sees it that way. Though Microsoft has won plaudits for successive editions of Halo and the Forza car-racing series, analysts have pointed to the titles’ relative age—Halo debuted in 2001; Forza first appeared four years later—as evidence that Microsoft’s homegrown studios have run out of ideas. “We have work to do there,” acknowledged Spencer, the Microsoft gaming chief. “We haven’t done our best work over the last few years with our first-party output.”
That must change if Microsoft, the only video game veteran among the Big Three consumer cloud companies, hopes to maintain its natural advantage against Amazon and Google. After all, in video games, as in other parts of the media industry, content is king—which is why Microsoft’s rivals have moved to hire gaming veterans from top shops such as Electronic Arts (Madden NFL, Need for Speed) and 2K Games (Civilization, NBA 2K20) in an effort to build their own franchises. It is an uncanny echo of the moves by Amazon and Google to build their own premium programming, for Prime and YouTube, respectively, to compete with Netflix.
But Rome wasn’t built in a day. Seven years after establishing a gaming group in 2012, Amazon laid off dozens of game developers as it reorganized itself for a cloud-based future. (Amazon downplayed the news. “Amazon is deeply committed to games and continues to invest heavily in Amazon Game Studios, Twitch, Twitch Prime, AWS, our retail businesses, and other areas within Amazon,” a spokesperson tells Fortune.)
Van Dreunen, the SuperData analyst, believes it will take up to five years before cloud-driven efforts by the Big Three will significantly affect the traditional gaming industry. Until then, look for cloud computing’s leaders to continue investing in their data center infrastructure to support the “gradual rollout” of cloud-gaming services, he says.
Why would Amazon, Google, and Microsoft make so much noise about a future that’s so far away? It’s all a part of the “land and expand” business model familiar to the technology industry, says analyst Pachter: Give a speech, plant a flag, hope that early momentum snowballs into an insurmountable competitive advantage. After all, “Facebook wasn’t a billion-dollar idea until it was,” he says. “Uber wasn’t a billion-dollar idea until it was.”
Microsoft, in particular, has no intention of missing out. The company still regrets losing the mobile war to Google and its Android operating system. (Microsoft “missed being the dominant mobile operating system by a very tiny amount,” cofounder Bill Gates lamented earlier this year.) To underperform in an area where it has a head start of almost two decades would be, in a word, unconscionable.
Time to suit up, then. “We’re in gaming for gaming’s sake,” Nadella says. “It’s not a means to some other end.”
A version of this article appears in the August 2019 issue of Fortune with the headline “Big Tech’s New Street Fight.”
More must-read stories from Fortune:
—Netflix isn’t in trouble without Friends. It just needs to work harder
—Could A24’s The Farewell be this summer’s biggest indie success story?
—Classic rock is reanimating Summer 2019’s movies
—How Marvel pictures the future post Spider-Man and Endgame
—Listen to our new audio briefing, Fortune 500 Daily
Follow Fortune on Flipboard to stay up-to-date on the latest news and analysis.
