Steam-powered ideas: a market for open innovation
Videogaming is a notoriously challenging industry. How then has Valve been able to develop such a dominant position since its founding in 1996?
It is difficult for publishers and developers to survive in the videogaming industry. Product development is technologically and creatively complex, projects can cost millions of dollars for a AAA title (AAA is the industry’s informal classification for videogames produced and distributed by mid-sized to large publishers, typically with substantial development and marketing budgets), timelines are notoriously unreliable, and even if the project is eventually released to the public, user tastes are fickle.
How then has Valve, a small Washington-based developer, grown into one of the world’s most successful and influential videogame companies? As of 2017, Valve accounts for an estimated 18% of the $24 billion worldwide digitally-distributed PC videogame market, serves 125m registered users on its Steam platform, and is considered to be the most profitable company in the USA on a per employee basis. Valve has achieved this success through its long-term embrace of open innovation as a fundamental element in its business model.
Open innovation, often also known as crowdsourcing, occurs when a company leverages external participants who contribute labour or ideas that could alternatively have been performed or generated internally by employees or contractors. These contributions can involve both idea generation (e.g. innovation tournaments such as Netflix’s $1m prize for a better recommendation algorithm) and also idea selection (e.g. Threadless’ use of user voting to narrow initial design submissions to a shortlist).
Open innovation has been critical to Valve’s success since it was founded. Many of Valve’s most successful videogames such as Counter-Strike and Team Fortress began as “mods” in the broader user community before Valve brought the IP and teams behind them in-house for development into fully-fledged retail products (an approach at odds to that adopted by the broader videogame industry).
As Valve matured with the launch of its Steam platform, its business model transitioned from identifying innovative ideas of others and refining them in-house to building an environment where others could develop content and make it available for sale, with Valve taking a 30% cut of all revenue generated on the Steam platform. Each feature introduced to Steam, be it the release of the Steamworks API, the creation of the Steam Workshop for the sale of user-generated content (most famously seen in Team Fortress 2’s “hat-based economy,” with Gabe Newell once commenting “…we have a kid in Kansas making $150,000 a year making [virtual] hats…”), or the introduction of user reviews, has been focussed on increasing the amount of content available for purchase and helping users identify great content to buy. In particular, Valve has focussed substantial efforts on reducing the difficulty for users to make content available for sale on Steam. While Valve initially curated all content launched on Steam, Steam Greenlight was launched in 2012 giving users the power to vote on what content they wanted to see launched on Steam. After user and developer criticisms, this program evolved into the 2017 launch of Steam Direct, a new model whereby developers need only fill out some simple identification and tax documentation, pay a refundable $100 deposit, and wait 30 days for the content to go through a minor review from the Valve team before their submission will be launched on Steam. The cumulative result of these various efforts can be seen in explosive growth in Steam’s user base, available content, and revenues.
But as Gabe Newell and the Valve team look to the future, they must be cautious. Valve’s success has been facilitated by an open PC ecosystem, but as players such as Microsoft and Sony push users into closed-platform environments, Valve will need to invest to sustain the open nature of PC videogaming. Indeed, Valve’s recent efforts to reinvigorate gaming on the Linux operating system and make investments in proprietary hardware can be seen as responses to those industry dynamics. Valve will also need to improve its systems for dealing with inappropriate content. Recent controversies have highlighted the tension between reducing barriers to publication and maintaining standards.
Whether Valve can successfully navigate these challenges will depend on their ability to answer two questions:
- How can Valve expand monetisation of user-generated content without disrupting the dynamics of communities that have evolved without monetary incentives?
- How can Valve prevent the spread of inappropriate material on its platform without stifling users’ ability to release content or incurring excessive monitoring costs?
 Taylor Soper, “Valve reveals Steam’s monthly active user count band game sales by region,” GeekWire.com, August 3, 2017, [https://www.geekwire.com/2017/valve-reveals-steams-monthly-active-user-count-game-sales-region/], accessed November 2018 and Dustin Bailey, “With $4.3 billion in sales, 2017 was Steam’s biggest year yet,” March 23, 2018, [https://www.pcgamesn.com/steam-revenue-2017], accessed November 2018.
 Andrew King and Karim R. Lahhani, “Using Open Innovation to Identify the Best Ideas,” MIT Sloan Management Review 55 (Fall 2013), 1.
 Valve Handbook for New Employees (Valve Press, 2012), PDF e-book, accessed November 2018.
 Dan Crabtree, “Gabe Newell: “Windows 8 is kind of a catastrophe” ,” IGN.com, July 25, 2012, [https://www.ign.com/articles/2012/07/26/gabe-newell-windows-8-is-a-catastrophe], accessed November 2018.
 Jeffrey Matulef, “Steam Greenlight to be replaced with Steam Direct next week,” Eurogamer.com, June 6, 2017, [https://www.eurogamer.net/articles/2017-06-06-steam-greenlight-to-be-replaced-with-steam-direct-next-week], accessed November 2018.
 See the recent controversies with titles such as Active Shooter, ISIS Simulator, and AIDS Simulator initially published and then removed from the Steam platform. Nat Levy, “Valve removes controversial school shooting game and denounces developer as ‘troll with history of customer abuse’ ,” GeekWire.com, May 29, 2018, [https://www.geekwire.com/2018/valve-removes-controversial-school-shooting-game-denounces-developer-troll-history-customer-abuse/], accessed November 2018.
Student comments on Steam-powered ideas: a market for open innovation
Steam is AWESOME!
One other thing made Steam successful is that they maintain high LTVs from users with their recommendation system. The recommendation system is the key to continuously generate revenue, especially for outdated or indie games – it is extremely difficult (or annoying) for the users to wandering the tons of games on the Steam database. Valve has been doing very well on recommending the appealing games to the users based on their past purchase, likes, searches etc. Even if the user doesn’t buy it immediately, once the user “likes” the game, he/she highly likely return to Steam to buy the title when Valve notifies that title is on discount through email.
But I’m a bit doubtful on monetizing user-generated content, specifically the mods – the resist from users would be huge, as they are accustomed to use mods in free. Indeed, Valve announced to sell some of the mods of Skylim 2015 but turned down the policy because of the strong opposition from the user community (https://www.polygon.com/2015/4/27/8505883/valve-removing-paid-mods-from-steam). (I remember they announced the premium mods of Fallout afterwards, but not sure about the results)
Instead, I think they could allow users to create their own DLCs (Downloadable Content) which have been created by the official publisher and additionally sold through steam. The downside is that DLCs are much complex to create than mods, so Valve might need to make a huge investment to foster more user-created DLCs and deal with the IP issue with original publisher and the DLC creators.
And it’s quite surprising that they continue facing the trolling even with the Steam Direct system – I thought the bars were high enough..
I’m a big fan as well of Steam’s leveraging of human capital without the commensurate overhead costs, but if I were to channel my inner Steve Jobs (or Henry Ford), one could argue that the fans rarely know what they want until we create it from them. In that sense how would you as management balance the incremental nature of crowd-sourced innovation vs. step function improvements driven by the company, but at higher cost? Moreover, besides the risks associated with inappropriate content, how can the company maintain a consistent brand / value proposition when they hand the narrative over to the public?
I am a big fan of Valve’s history in tapping into its communities to develop user-generated content, but am unconvinced that monetizing this user-generated content will be fruitful. User engagement keeps customers immersed in the game, but should not be exploited. Several video game companies have done this successful to date. Activision Blizzard, for instance, has a game called Overwatch, which has exploded in growth as of late, partially due to its commitment to listening to its customers. Blizzard actively sources the input of its users to make changes (sometimes dramatic), which fundamentally change the game itself. In other words, the game is a constant work-in-progress, feeding in the collective input of its users, to dynamically shape the game to the delight of its customers. With its captive customer base, it then later monetizes through more traditional means (e.g. in-game purchases). I believe Valve can take a page out of Activision’s book by creating this kind of iterative model, versus treating the problem as a dichotomy of pure user-based input or no user-based input.
Thanks for your article, it adds another interesting perspective to Valve over what we learned in our Valve case. You paint an interesting trend on the tension between Steam’s open innovation and the direction gaming consoles are advancing to.
For your second question – how to prevent inappropriate material – I believe the answer can be combined with another megatrend, machine learning. Policing an open-content platform is a task that is not unique to Valve. For example, Instagram uses an ML algorithm for anti-bullying: https://www.techrepublic.com/article/how-ai-became-instagrams-weapon-of-choice-in-the-war-on-cyberbullying/.
Implementing such a system can save Valve precious manpower and allow users to keep creating content freely.
Hi Oliver – great article. Valve is such an interesting company and I think has really transformed the video-gaming space with Steam and it’s open platform. I completely agree with Haerim’s point about the hesitation of the masses to pay for mods when they are already used to having them for free – you risk backlash and potential loss of many customers this way. I’m not sure if there’s a nice solution, but I do think the practice of hiring extremely popular/talented mod developers is an option – this might be useful and be ‘crowd’ acceptable. Just a thought!
Thanks for this very interesting article. I think Valve is a pretty interesting company in terms of its leadership philosophy and it was fascinating to learn how efficiently profitable it was (at least in terms of human capital). Another interesting strategic question related to open innovation is one that is common to many ‘platform’ companies. Valve is earning a lucrative 30% cut for all games sold on their Steam platform. Will these margins compress over time? If not, will there be another platform that develops where game developers can keep a higher percentage of their earnings. We have seen such a competition develop between Uber and Lyft, as well as many other smaller competitors. In the video game context, we have further competition from closed platform, such as Microsoft and Sony. I’m curious to see how Valve’s fortunes change over the next 5-10 years as successful developers may look for either closed platforms or new forms of open innovation platforms.
I believe Valve has a moral obligation to society to prevent inappropriate material from being published on its platform. Video games are becoming more and more popular, not only in the context of people playing them, but also the emergence of E Sports. To help protect customers from inappropriate material Valve should look into practices used by social media firms to monitor for inappropriate postings and then flag or remove accounts that post such material. If the postings are illegal in nature then Valve should work with law enforcement to help identify nefarious actors.
Thank you very much for the article, Oliver. It is indeed very interesting how Valve was able to build their success leveraging the ideas of its (very) creative customer base. I also agree that opening the creative process to customers might generate some inappropriate content, but I believe that Valve can offset that risk by taking two measures. First, I believe Valve could create/purchase machine learning to understand better which content might be offensive (e.g. searching through comments on that game within the platform). Second, I believe that Valve could share this responsibility with users, by creating a very clear and easy to use channel for players to whistle-blow issues related to content to the company.
Very interesting read! One way to potentially increase monitoring of the content while keeping costs manageable is to leverage machine learning solutions. Filtering spam email for instance is one of the oldest and most successful commercial use cases of machine learning, and a more elaborate version could be applied to Steam, whereby a machine learning algorithm would review the content proposed first, and either publish it, reject it, or push it to a human to double check if the confidence level of the algorithm is below a certain threshold.
While I am generally a huge fan of Valve and their attempt to democratize game development, I am somewhat weary of their attempt at higher monetization, specifically their culture of extreme discounts. Steam has trained their user-base to wait for season sales and expect deep discounts on any type of game (not only classics or other cash cow equivalents). This further stifles the gamers lack of appreciation for content creators and, in the long run, hurts the community of independent devs. To provide an anecdote: A good friend studied game development at UCLA and is a avid producer for indie games. His ability to push sales on full priced games is equivalent to zero. His ‘personalized recommendation’ from the Steam customer success team is to apply 50% or more discounts to games that have an original price tag of $10. I am worried that steam perpetuated a culture of price dumping that in the long run will hurt the independent developer community.
Oliver, I really enjoyed this deeper dive into Valve’s approach to open innovation, particularly the Steam API. It certainly makes sense that as the hardware landscape changes, Valve will need to continue to push for PC gaming in comparison to Sony’s Playstation and Microsoft’s Xbox. Given their prior success in using open innovation to drive new game ideas and “mods”, I wonder if Valve should fund a similar effort in hardware? Can open innovation help Valve encourage PC gamers and development of gaming hardware with the PC? Should Valve open the specs of their existing controller to allow enthusiasts to tinker and improve upon their hardware?
Very interesting, thanks Oliver. Valve seems to have a pretty shrewd business model in terms of making money off of others (why do the work yourself, when you can have your customers do it instead). I definitely worry about the seeming lack of controls / self-policing community and risk of inappropriate content – Valve should have a separate task force monitoring this at all times. Of course, the tradeoff would be the potential alienation of several developers and gamers.