435-015 ETHAN BERNSTEIN FRANCESCA GINO BRADLEY STAATS HARVARD BUSINESS SCHOOL 9-415-015 BET: AUGUST 10, 2013 Authorized for use only in the course MGMT 6125 Perspectives for Business Analysis at Fanshawe College taught by Robert Brookes from Jan 06, 2020 to Apr 30, 2020. Use aubide these parameters is a copyright violation Opening the Valve: From Software to Hardware (A) Manager-The kind of people we don't have any of. So if you see one, tell somebody, because it's probably the ghost of whoever was in this building before us. Valve Handbook for New Employees After dropping out of Harvard in 1983 to join Microsoft, Gabe Newell, co-founder of Valve Software, spent the next 13 years as "producer" for the first three releases of Windows, becoming a Microsoft millionaire in the process. Cabe explained: Like a lot of people in my cohort at Microsoft, I hit the lottery-I reached the point financially where I could do anything. I could retire and sail around the world. But the thing I enjoyed the most was working with smart people who liked building products that affected lots of people. That's what I'd gotten hooked on at Microsoft, and that's what I created Valve to do. So the question was why would Valve be an interesting company? We had to think both about the design of our products and the design of the company. By offering unheard of levels of autonomy, Gabe intended to attract talent and allow it to flourish. By choosing an industry in which flourishing talent translated into high returns, he hoped to afford commensurately high pay and rewards. As Gabe looked out on the horizon in 1996, the video games industry (see Exhibits 1 and 2 for a short history) seemed poised to offer the increasing returns to talent required to power such a model. Therefore, he and co-founder Mike Harrington, with experience in operating systems but not in video games, created Valve Software. Valve was established as a radically open, non-hierarchical company where job titles and org charts were nonexistent, employee time was 100% self-allocated, and every desk was on wheels so that employees could physically relocate themselves any time they wished. By mid-2013, Valve's unmanaged 400 employees had created many of PC gaming's leading titles, supported by Valve's digital distribution system, "Steam," which helped transform the business of selling video games. But Valve saw a plateau ahead. Through Steam, Valve's games could be played on PC, Mac, and Linux-but without accessible living room and mobile hardware, Valve's growth opportunities were limited. Innovations in hardware would be essential to Valve's ongoing success. Could a "boss-free" company designed to build gaming software create hardware too? HS Prod that and Francesca Gine and Professor Badley Staat (UNC Kenan-Ragle Business School prepared this case with the assistance of Research Associate Luke Hall it was viewed and approved before publication by a company designate Funding for the development of this case was provided by Harvard Business School and not by the company cases a developed solely as the basis for da discussions are not intended to serve as endement, serves of primary data, or illustrations of effective infective Copyright 2004, 2015 President and Follows of Harvard College. Tender copies or request penision to produce materials, call 1-800-55 Business School Publishing, Boston, MA (216), or go to www.hipharandada. This publication may not be digitized, photocopied, wie produced, posted, or transmitted, without the permission of Harvard Business School Opening the Valve: From Software to Hardware (A) The Creation of Valve We were convinced in 1996 that video games had an inherent advantage over passive, couch-potato entertainment: consumers could participate in the experience. We could point to many psychological studies as to why agency makes video games a more valuable experience. Gabe Newell, co-founder of Valve By 2012, consumers spent an estimated $20.77 billion on video games worldwide, or roughly two- thirds of the global movie box office sales.' Fifty-eight percent of Americans played video games, in one medium or another, and 51% of US households owned at least one dedicated game console. Major releases, known in the industry as AAA (triple-A) titles, could attract millions of concurrent users and regularly competed with more passive entertainment forms (like movies) for audiences. Looking back to Valve's founding in 1996, Gabe recalled that they had a specific strategy in mind: We were pretty sure that video games were going to continue to increase as a percentage of total entertainment spend. We were also pretty sure that making video asmac wise had that the docina al vios armse sore sad still in folk set and choosin aught by Robert Brookes from Jan 06, 2020 to Apr 30, 2020. 2 415-015 Opening the Valve: From Software to Hardware (A) The Creation of Valve We were convinced in 1996 that video games had an inherent advantage over passive, couch-potato entertainment: consumers could participate in the experience. We could point to many psychological studies as to why agency makes video games a more valuable experience. Gabe Newell, co-founder of Valve By 2012, consumers spent an estimated $20.77 billion on video games worldwide, or roughly two- thirds of the global movie box office sales. Fifty-eight percent of Americans played video games, in one medium or another, and 51% of U.S. households owned at least one dedicated game console. Major releases, known in the industry as AAA (triple-A) titles, could attract millions of concurrent users and regularly competed with more passive entertainment forms (like movies) for audiences. Looking back to Valve's founding in 1996, Gabe recalled that they had a specific strategy in mind: We were pretty sure that video games were going to continue to increase as a percentage of total entertainment spend. We were also pretty sure that making video games was hard-that the design of video games was, and still is, folk art and changing very rapidly. Most projects would fail, but a few blockbusters would be extremely profitable. The question was: are blockbusters randomly distributed, in which case you just want to have a lot of irons in the fire, cast your net as wide as you can, and then hope that one will hit? We didn't think so. To us, there was something at work other than just random chance. It seemed certain people who were successful creating a blockbuster game tended to be consistently successful. Predictable success was all about attracting and retaining the specific individuals on each team who seemed to predict success. But it was not just about recruiting stars. One of the lessons we had taken away from Microsoft was that no matter how good you are as a set of individuals, after you've worked together through multiple shipping iterations, you're adding more value to the underlying capabilities of the people on the team. The more stable you can keep those high-performing teams, the better. To this day, we're startled by how Hollywood creates and destroys their production teams multiple times over the course of making a film, which to us just seems like madness when we know that working together only improves over time. So the key question was how to find these incredibly talented people, convince them to come, and then convince them to stay together over time. The answer to that question was written in the Valve Handbook for New Employees (Handbook), which was released to the public in 2012. Spontaneous Order: Valve's Amorphous Corporate Structure Michael Abrash, a fellow Microsoft veteran whom Gabe had been trying to recruit for years and who finally joined Valve in 2011, recounted how Gabe set about organizing Valve: When Gabe was at Microsoft in the early 1990s, he commissioned a study to find out what was on customers' hard disks; Windows was number two and Doom was number one. The idea that a 10-person company of 20-somethings in Mesquite, Texas, could have gotten software distributed more widely than the biggest software company in the world made him think that the world had changed. So he did some research. He found Authorized for use only in the course MGMT 6125 Perspectives for Business Analysis at Fanshawe College taught by Robert Brookes from Jan 06, 2020 to Apr 30, 2020. Use outside these parameters is a copyright violation. Opening the Valve: From Software to Hardware (A) 415-015 415-015 Opening the Valve: From Software to Hardware (A) company vacation. Valve also tried to engage its employees' families, believing that, Gabe noted, "The company does not hire people; it actually hires their families." Gabe believed that the key aspect of Valve's ability to retain its employees, however, was the system itself. Old Management's Antiques: Stack Ranking Although it eschewed most hierarchical structures, Valve stack-ranked its employees for compensation purposes-but all ranking was done by peers. Employees were first ranked internally within cabals by those with whom they chose to work, based on their technical ability, productivity, contribution to the operation of the group, and the successful development of the resulting product. The cabal peer rankings were then aggregated and calibrated by committees of peers across cabals to produce a company-wide ranking upon which compensation was based. Valve's Path Most of the famous names in video game production-Sony (PlayStation), Nintendo, Microsoft (Xbox), Sega, Atari, and Activision-had been built on a classic razor-razorblade business model: they sold low-margin video game consoles that connected directly to the household TV, in order to sell compatible high-margin games. Console operating systems were closed, lest third parties produce competing high-margin games that ran on a manufacturer's low-margin consoles. Industry standards required that third parties that wished to develop games for another company's consoles pay a royalty to the console manufacturer for every compatible game sold. (See Exhibit 1 for a short history of video games.) But by the time Valve was established in Kirkland, Washington, in 1996, the world had begun to look very different. Households increasingly owned a PC, and over 80% of PCs-regardless of manufacturer-ran a Microsoft operating system (OS), which was not only ubiquitous but also open to developers. As long as Valve could use that open ecosystem for its games, prospective customers only had to be convinced to buy the game, not an expensive (yet low-margin) machine to play it, and Valve could seemingly ignore hardware altogether. Half Life and Mods: Building an Open Community One of Valve's first acts was to acquire a license to the Quake game engine. This enabled Valve to make Half-Life. The game combined suspenseful combat with puzzle solving. The game won over 50 game-of-the-year awards and went on to sell 9.3 million copies. In 1998, Valve hired the developers of the popular game Team Fortress, who then created an updated version for Valve." Valve decided to push the "open" model one step further. Traditional video game companies were known to closely guard their IP, fearing that a failure to do so would lead to severe brand damage. Yet with the spread of the Internet, there was a growing market for online-distributed modifications ("mods") for existing games-essentially hacked versions of existing games. In a departure from the mainstream, Valve embraced that trend and even used it as a recruiting technique. In 1999, college students Minh "Gooseman" Lee and Jess "Cliffe" Cliff developed the Counter-Strike mod for Half-Life, which transformed the game into a multiplayer arena where teams of players fought over a fixed map to complete mission objectives. Rather than enter an IP battle, Valve opted to hire the two, bringing Counter-Strike development in-house in the process. A full retail version was released in November 2000. Counter-Strike was the first step in establishing a pattern at Valve of crowdsourcing mod development and bringing particularly notable mods in-house along with their creators. From First-Person Shooter (FPS) to Digital Download: The Rise of Steam In 2002, Valve announced the release of their Steam platform. Steam's original purpose was to form an integrated system for auto-delivering and installing software patches for Valve's games, especially online multiplayer games like Counter-Strike, which required all players to be running the latest patch. Valve also announced that it would use the Steam platform to make a variety of the tools it had created available to outside developers and "modders."10 Valve began using Steam to retail its own games in 2004, opened the platform to other games in 2005, and used it to embrace several emerging trends, including "free-to-play" and digital/mobile distribution (see Exhibit 3). Steam rapidly dominated digital distribution of PC games. Although Valve did not disclose exact numbers, industry observers believed that 50%-70% of the PC digital distribution market in 2012 ran through Steam." As of 2013, the platform offered over 2,000 games, had 75 million active user accounts, and had set a new record of over 6 million concurrent users online.12 The Source of All Things Having utilized the Quake engine to develop its early games, Valve eventually sought to bring design entirely in-house. The result was the Source Engine. The engine enabled Valve not only to deliver superior graphics, but also to increase the modularity of development and thus the capacity for rapid adaptation. The first game to be made with the new engine was Half-Life 2, which was very positively received, selling over 12 million copies by 2011 while winning numerous awards.13 While the first use of the new engine was a sequel, Valve had broader ambitions to use the engine to build novel gaming experiences. Chief among the results of this strategy was Portal, a game that shared the camera perspective and control schema of Half-Life but required the player to kill only once and at great cost.14 The player's "gun" created portals, allowing the solving of puzzles and navigation around obstacles. The game was also known for its humor. Other games created using the Source engine included zombie survival game Left 4 Dead, the long-awaited Team Fortress 2, and third-party titles like the gothic role-playing game (RPG) Vampire: The Masquerade-Bloodlines. Valve versus the Brave New World Having benefitted greatly from the PC ecosystem in which it had grown up, Valve was increasingly striving to move beyond it. Valve released a PlayStation 3-compatible version of Steam (with its retail features inoperable as mandated by Sony) that allowed players of Valve games such as Portal 2 to access extra content and other Steam features. 15 Meanwhile, Valve continued to add functionality to Steam, including a streamlined support system for mods and the capability to allow users to directly support other users of Indie games. Steam Greenlight allowed smaller developers seeking to sell their games on Steam to post information, video, and screenshots of their games and build community support. 16 Steam users could then vote on whether or not they wanted a game to be available on Steam. The Steam Workshop leveraged Steam's existing systems to allow for the centralized uploading, promotion, and dissemination of mods or other user-created content for games. While the rules differed by title, Steam made these customizations easily accessible to PC gamers unwilling to perform complicated installations and configurations.17 As it opened its own platforms, Valve carefully monitored the growth of closed-shop platforms. In 2011, top PC game publisher Electronics Arts ("EA") established its own digital distribution platform 6 is a copyright violation. Opening the Valve: From Software to Hardware (A) 415-015 415-015 Opening the Valve: From Software to Hardware (A) called Origin for EA-developed games and was flirting with making its games exclusive to Origin. 18 When Microsoft released Windows 8 with an integrated, closed Windows Store, Gabe vocally expressed concern that it could decrease the open character of PC gaming. 19 In part in response to these developments, he stated that Valve would deepen its involvement with the Linux platform as a hedge against similar future developments.20 Toward Hardware? Any decrease in the open architecture of PCs was concerning to Valve. Coomer explained that Valve's "success story as a medium-sized company is entirely based on the existence of that openness in PCs." Even so, open architecture had limitations since Valve had no control over the hardware on which its games were running. Abrash echoed a common frustration at Valve when he noted that "Valve was increasingly concerned that taking advantage of advanced opportunities in software can't happen until the hardware changes." Meanwhile, Valve viewed market data such as that in Exhibit 4 as a call to action by its customers. So, spontaneously yet slowly, a cabal emerged at Valve committed to exploring hardware. As one of its members explained: A group of us started to believe that it was important to actually respond to customers' requests about not having to abandon either their games or their friends because each was in a different room. We started to scope the work required to actually connect the dots for customers to help them get the games out of their computer stations and into the living room. As a first step, in 2013, Valve developed Big Picture, which adapted the Steam platform for use on wide-screen TVs.2 Big Picture was conceived as part of a general strategy of integrating PCs into the living room and allowing PC gaming to compete more directly with consoles. The team behind it understood that Big Picture itself could not accomplish this goal. Instead, Big Picture was seen as both a smaller, software-based first step in that direction as well as an effort to generate internal credibility for the project and support the accretion of resources for further work. One of the early cabal members, Anna Sweet, commented: I think it became clear to everybody that the other pieces-input devices and dedicated console-like hardware-needed to be developed by Valve in order to solve our customers' problem of wanting to move into the living room. Building internal and external support for the project was particularly important because the PC had failed repeatedly to make its debut in the living room, and there was a risk that Valve's effort could meet a similar end or at least be tarred by that memory. The hardware cabal was tiny at first, but then, like other successful Valve projects, it accreted people and effort and arms for the work. The cabal began to attract software engineers who were accomplished enough hardware hobbyists to get basic work done. They experimented with things like biometrics-collecting data about arousal state in the user's body through sensors that measured heart rate, blood volume, skin galvanic response, etc.-along with eye tracking and, to a limited extent, virtual reality. They built "Frankenstein" prototypes in-house, cobbled together out of low- tech solutions but sufficient to gather data and iterate. One hardware cabal member reflected, "In hindsight, we were just competent enough to be dangerous, but we did learn a fair amount." 0 to Apr 30, 2020. To produce higher-fidelity prototypes that could be used in robust gameplay testing, the cabal enlisted the help of outside contract firms, including a local firm full of experts in industrial design, electrical engineering, and manufacturing. For example, that firm helped Valve create a mouse that was covered in metallic contact points to collect skin galvanic electric responses with which the user's arousal state could be inferred and correlated with game events. With that mouse, Valve conducted game-design experiments to see if changes to game design could meaningfully affect arousal state of an individual playing games like Left 4 Dead. Yet even as the cabal grew, it also became increasingly clear that entering the hardware field could be challenging for Valve. These early experiments were rich with learning but also, as one cabal member put it, "messy," "unreliable," and "glacially slow." Valve was also "uncomfortable with outsourcing product development to outside contract firms, having long ago come to believe that it's exactly the kind of work which ultimately must be done in-house." There was a growing view that, if Valve was going to make a meaningful dent in the hardware space, they needed to make a significant commitment to hire hardware experts as full-time employees. But the nature of hardware development posed a potential disconnect with Valve's open staffing and T-shaped skill preferences. Abrash commented, "Hardware tends to be more an interaction of a team of specialists whose skills interlock, and generally requires far more planning and upfront investment." Coomer noted that producing hardware required the writing of "really large checks that other parts of the organization don't need to write." And perhaps most importantly, hardware production was marked by long lead times, requiring specifications to be fixed long in advance of release. Coomer reflected, "In our model, it's really hard for us to have things fixed. We don't think that way. Why would we stop designing if we can continue to improve it? But in hardware, it's much harder to have something be fluid." Valve had begun thinking about hardware in an organic way, but the next steps demanded far more deliberate planning and constrained activity. While irregular corporate organization was not necessarily unusual in software firms, would the Valve operating system be compatible with hardware production? Does Anybody Have a Walkthrough"? Valve faced three key decisions as it considered entering the hardware field. First, was entry into hardware feasible for Valve, given its unique structure and the competing demands of hardware development? For example, Coomer observed that "one of the most difficult challenges for us is that in the software model, what we like to do is put something out there and get real-time feedback and then constantly iterate on it," but putting an unfinished hardware product into the hands of consumers was not possible with the long lead times, up-front investments, legal and regulatory challenges, and complex ecosystems involved in producing hardware. Second, if Valve did step into the hardware segment, how should it do so? Should Valve mass- manufacture its own input and console devices for sale to end consumers? There were certainly those at Valve who found the idea appealing. Frank Taylor explained, "It's an interesting question because I think we have a lot of faith that, for the things we choose to do, we can do a better job than almost any other organization." But such a major intervention could potentially strain Valve's flat structure. Given Valve's approach to acquiring and motivating talent, and getting its employees to collaborate "In PC gaming, a "walkthrough" was a series of steps (solution manual) through which a level or game could be solved. Authorized for use only in Authorized for Opening the Valve: From Software to Hardware (A) and coordinate, was Valve really set up to manufacture hardware? For example, would Valve need to create a separate division that could operate differently in order to produce hardware? Others in the hardware cabal wondered whether it might make sense for Valve to instead adopt a more hands-off approach, working with partners and external parties. One member reflected: Perhaps Valve should engage in hardware as a company that exclusively innovates and produces prototypes and reference designs, which other companies take as guidance from us and essentially become engaged in the production process using the invention and innovation that we have pushed forward. But if Valve decided to outsource the manufacture of hardware to partners and external parties, it would need to carve out a productive role within the existing, interdependent hardware ecosystem- from retailers to component manufacturers to existing PC manufacturers and the software content producers already working with Valve via Steam (see Exhibit 5 for a visual map of the ecosystem). Valve was uncertain how to choose among potential partners and uncomfortable being too dependent on any of them. There was also a growing concern that partnering closely with outside organizations, all of which operated as more traditional, bureaucratic structures, might result in more (vs. less) pressure on Valve's unique structure. Third, should Valve adopt the same or different approaches to manufacturing the various components necessary to bring PC gaming into the living room? In addition to a console, users were also demanding a new controller to serve the wide variety of PC games, many of which had not traditionally been well-served by existing options. The plethora of competing factors and questions made it all the more difficult for the hardware cabal as they wondered how Valve could best participate in pushing hardware forward, and how it could do so without compromising its unique organizational structure. Use outside these parameters is a copyright oa taught by Robert Brookes from Jan 06, 2020 to Apr 30, 2020. Exhibit 1 A Short History of Video Games VIDEOGAME PREHISTORY 1947-Thomas Goldsmith and responsible for Authorized for use 1975-1977- The Early PC- 10861 415-015 415-015-10- (Nintendo) GAME OVER 1983-Just six years after aders for home ACTIVISION GAME OVER weave to form a third crash hit market leader Atari hard. SEGA MICROSOFT 1985-1989 ccess, Japanese a 1985- bankruptcy and e mainframe. The first PO Gaming Emerges a company with friend Steve From their earliest days, PCs were a platform for