The Promise of Cloud-Native Games

First published at a16z.com

Much of the buzz around cloud gaming has centered on its potential to “kill the console”—to remove the need for local hardware to play games. But the ongoing focus on hardware fails to grasp cloud gaming’s true potential. The real innovation in cloud gaming won’t simply be in how we play, but what we play: “cloud-native” games will completely upend the gameplay experience itself, as well as how those games are marketed and sold.

In cloud gaming, all processing and storage is conducted on servers in cloud data centers; the resulting graphics are streamed to player displays. It’s a model similar to Netflix streaming, but much more complex, since games require real-time input and low latency. Proponents believe that cloud gaming will dramatically expand the market for games, attracting consumers new to the industry (including those previously discouraged by the price tag of consoles or gaming PCs) and allowing users to play their favorite titles on-the-go.

In November, Google launched its Stadia cloud gaming service in beta. Competitors aren’t far behind: Microsoft is expected to unveil its xCloud service this year, and Amazon recently announced Wavelength, a technology for 5G games streaming.

That’s not to say the revolution is imminent. The hurdles are formidable, from minimizing lag to building out content libraries to figuring out new pricing models. OnLive, the first major attempt at cloud gaming, was introduced in 2010 and struggled with technical difficulties before unceremoniously shutting down. Its successor, Gaikai, was folded into Sony Playstation and struggled to build an audience. Still, cloud gaming’s breakout moment may come sooner than you think.

What’s different today? For one, better hardware, such as greater broadband penetration, improved cloud coverage, and 5G networks. But the real key is software—the increasing integration of the cloud in modern games. Some of the largest games today, from Epic’s Fortnite to Supercell’s Clash of Clans, run most of their network in the cloud. Even single-player games rely on the cloud for web services like account management, commerce, and analytics. Platforms like Twitch and Huya have also taught millions of users how to engage with live, cloud-streamed video. Though they’ve stopped short of true cloud streaming, games like Fortnite are essentially first-generation cloud games.

The logical progression of this trend will be “cloud-natives,” games written specifically for the cloud, in which client and server are hosted in the same architecture. This next generation has the potential to yield entirely new gameplay experiences and business models. These native games—exclusive to and solely playable within the cloud—will be the ultimate driver of cloud gaming. The following are just a few ways cloud-native games are likely to revolutionize the entertainment industry.

Cloud-native MMOs will grow like social networks

Most MMOs (massive multiplayer online games) have inherent network effects, meaning the games are more fun to play with more players. Yet MMOs often suffer from a “cold start” problem: In the beginning, there aren’t enough players to create a positive experience, leading to churn for new users. Cooperative play with friends is the best way to onboard and retain new users, but there can be many barriers along the way. For example, users may play at different times or on different platforms. Finding friends in-game can be an ordeal, thanks to opaque matchmaking rules and server restrictions. (See an example of a matchmaking lobby from Overwatch below.) And if you play less frequently than your friends, you may enter a game at a different level, rank, or location.

Cloud-native MMOs (CMMOs) remove most of these barriers. Since they are inherently cross-platform, these games are accessible from any connected device. They have no download, installation, or load times, eliminating the need to wait three hours for a patch or to buy a physical copy of a game.

To ease onboarding, cloud games can use deep links that seamlessly allow new players to join a friend’s game session. While deep links that open apps are not new, a deep link to a cloud instance could drop a player directly into a game world amid their friends. Furthermore, these links could be customized to create different onboarding experiences based on the referrer’s group, guild, or faction. This form of personalized onboarding could render high-friction systems, like lobbies and matchmaking, obsolete.

Finally, users who want a more lean-back experience could watch friends on a cloud-delivered stream—imagine a real-time, interactive Twitch feed—and choose the exact point to jump into the action or assist as a spectator. Spectators could scout ahead for enemies or utilize a menu of helpful in-game abilities, such as gifting power-ups to players (similar to Sponsor Gifts in The Hunger Games).

Collectively, these cloud-enabled features accelerate the network effects inherent in multiplayer games. If successful, the first CMMO could flourish entirely through player-led recruitment, with a viral growth curve that more closely resembles that of Facebook than a traditional MMO.

Cloud games will create a new class of video marketers

Besides greater virality, cloud-native games will also enable new forms of marketing for AAA games, which have traditionally relied on retail marketing such as billboards and display ads. With no install times, prospective buyers will be able to click a link to immediately try a game—a huge advance.

“Click to play” will inevitably lead to video as the preferred medium in which to sell games. A 2018 survey found that 85 percent of US millennials reported buying a product after watching a video. Video ads leading to demos (“play an hour free”) are likely to grow in popularity, with games optimized to convert players within the first hour.

Video narratives are particularly well-suited to influencers on social platforms like Twitch, Tik Tok, and Youtube. With cloud games, creators will evolve from providing discovery to community (“click to join”) and direct commerce (“click to keep playing”). In the Google Stadia demo below, users clicked to join a live-streamed NBA 2K game:

As a leading indicator for the potential of “click to join,” Bixin, an emerging Chinese platform, has already grown to over 20 million users by matchmaking players with gaming companions. On Bixin, users pay top creators $8/hour for help learning new games, training for eSports competitions, or just hanging out.

As cloud games become ubiquitous, video and influencer marketing will become increasingly important. Over time, sales commissions and “click to join” may become the largest sources of influencer revenue in the cloud gaming economy.

Cloud games will enable AI-powered, real-time content generation

Because the client and server are in the same network, cloud-native games can track and collect data on almost every aspect of a user’s journey. This trove, conveniently stored in data warehouses in the cloud, can empower AI and machine learning in games in groundbreaking new ways.

For example, games have long monetized through the sale of cosmetics that alter a player’s appearance or the world around them. Because the cloud offers unlimited data, processing power, and minimal client-server latency, fully dynamic characters and environments can be generated by AI in real-time. The following clip of an Nvidia deep learning-based system shows a user modifying a photo-realistic, virtual environment with the help of AI:

In the future, real-time content generation could fuel new, immersive story-telling methods. The next generation of “choose-your-own-adventure” might be a virtual world that adapts in real-time to your choices. And to monetize these virtual worlds, personalized, spontaneous ads may emerge, similar to the biometric ads in Minority Report.

Over time, AI-powered, procedurally-generated worlds could provide an endless playground for users, not far removed from the vision of the Metaverse or the OASIS from Ready Player One.

These are only some of the potential benefits of cloud-native games. The original developers of early SaaS products like Salesforce or Gmail could hardly have envisioned the subsequent innovation that would spawn modern cloud ecosystems like Amazon Alexa or the Tesla Autopilot. Similarly, we are only at the beginning of the cloud gaming software era.

The outlook for cloud gaming

If history is a guide, we expect to see the first cloud-native games hit the market in two to three years. By comparison, the top games of past console generations usually launched in a similar time-frame—for example, the groundbreaking 3D games Gran Turismo and Final Fantasy VII were introduced three years after the initial Playstation launch; Halo 2, the first game with online matchmaking, arrived two years after the Xbox’s initial release.

Cloud-native games are coming—and software will be cloud gaming’s killer app. Spurred by investment from Google, Microsoft, Amazon, and many others, the next generation of cloud-native games will have the potential to reinvent the gaming experience as we know it.