Valve is a fascinating example of a company experimenting with a new way of working – one in which there are no traditional managers, and where employees are encouraged to take direct responsibility for choosing their own projects and completing them in an efficient and effective way. Of course, Valve is not alone in pushing a “manager free” model. Zappos (owned by Amazon) is currently experimenting with a similar model they call “holacracy”, and moving further back in time there are such classic examples as W.L. Gore and Associates’ “lattice structure”, Oticon’s “spaghetti organization”, and Brazilian industrial products company, Semco. Nonetheless, Valve’s experiments are noteworthy because they are ambitious, visible, and so far very successful.
Let me offer three comments about Valve’s Way, one practical, the other two more academic in nature. Practically speaking, Valve’s unusual model works in large part because the company is still relatively small. Puranam and Håkonsson (2015) point out some of the risks of Valve’s management model (people focusing effort on the wrong projects, duplication of effort, informal hierarchies emerging that marginalize some people, and slow consensus-building). I would expect these risks to compound with growth, and at some point (perhaps it has already been reached) the risks will outweigh the benefits, at which point the solution is to impose some sort of formal structure. For example, the company could divide itself – amoeba-like – into two or three separate companies, or it could create some low-touch processes for reviewing projects, hiring new people, or resolving disputes. This would be part of the natural evolution of the company and would help it to continue growing. Without such changes, Valve’s most senior people will find themselves spending all their time resolving disputes, fighting fires, and trying (perhaps vainly) to create coherence in the company’s offerings.
From an academic perspective, Valve is valuable to the research community because it allows us to see what might be possible. We all have our pet theories about how the world works, and the best way of verifying or falsifying them is when companies “stress test” them for us. For example, hierarchy is a basic principle of organizing, but depending on your viewpoint hierarchy might be a mode of coordinating economic activity, a means of exerting control over workers, or a fundamental attribute of any social structure. By building a company without a formal hierarchy, Valve provides evidence to enrich our theories. It seems likely that some sort of informal hierarchy is taking shape in the company, in the absence of a formal one. It also seems that alternative modes of coordination, based on mutual adjustment, are emerging in place of the traditional top-down mode.
A related point is that Valve, as a gaming company growing up in the Internet era, has access to knowledge-sharing technologies that earlier management pioneers such as Semco and Oticon could only dream of. I don’t know enough about the specifics of this case, but it seems likely that it is easier to coordinate activities in a bottom-up way today than it was twenty years ago. If informally run organizations used to hit their limits at around 100 employees back then, perhaps that limit has now risen to 200 or 300 people? This would be an interesting proposition to test. However, Valve also reminds us that there are limits to the Julian Birkinshaw What Lessons Should We Learn From Valve’s Innovative Management Model? 9 power of technology in organizations. Take the case of hiring, for example, which has been a challenge at Valve. Information technology might speed up the initial screening process, but the key parts of the hiring process cannot be done without high levels of human interaction.
The second theoretical issue I would like to note is about the rhetorical aspects of Valve’s new way of working, and how such ideas spread. Valve has received a lot of publicity for its manager-less model, in large part because journalists, consultants, and academics are always on the lookout for novel and progressive practices. This is a good thing, for the most part, but it is not without its problems. For example, it is too early to say for sure whether Valve’s model actually works: perhaps the company is performing well despite its model; perhaps its employees are behaving unusually well because they are in the public eye (a large-scale Hawthorne Experiment); perhaps the charisma of CEO Gabe Newell is garnering as much attention as company performance. As researchers, we should be cautious in our attributions. A separate concern is that we probably do not agree what is novel about the Valve model. The company is a Rorschach test for the research community: each of us looks at it through our own lens and sees what we want to see. I see Valve as a story about management innovation; others might focus on organization design, incentive systems, governance, information technology, employee engagement, and so on.
One could argue that this is not a problem, as any organization, by its nature, can be viewed through multiple lenses, all equally valid. But the real problem is that these lenses are often applied on the basis of limited evidence. Most of us have only second-hand information about what Valve is up to, and the net result is likely to be a disconnect between rhetoric and reality. Arguably, this is what has happened to celebrated cases like Semco, Oticon, Morning Star, and W.L. Gore and Associates. Students and readers of the Harvard Business Review become so enchanted with the hype around these companies that they end up disillusioned when they encounter these companies in reality or when company performance dips. And this, in turn, is unhelpful to the research community, as it reduces the legitimacy of our theories.
These risks cannot be avoided entirely, but they can certainly be mitigated by careful analysis and reflection. One simple rule of thumb is that we should all conduct some primary data collection – a couple of interviews, for example – before volunteering a theory about an unusual company such as Valve. This approach should ensure that our ideas do not get too far out of line from the real-life phenomena we are purporting to study.
References
- Puranam P, Håkonsson DD. 2015. Valve’s way. Journal of Organization Design 4(2): 2-4.
This article is governed by:
