Bossless or With a Boss? It's a Matter of Balance

Over the years, several companies have adopted radically decentralized, "bossless" organizing forms. Here, employees themselves initiate new projects, or join existing ones without interference from management (also termed as self-selection), rather than managers telling them which projects they should work on. Bossless organizing is not a fringe phenomenon: tech firms such as Valve Software and Github, have fully implemented it; and well-established firms, mature firms such as Google or 3M, have implemented it in some form. Self-selection is also found in other, more traditional industries, like in W.L. Gore (makers of Goretex fabric) and Morning Star (a food processing company).
But it seems that there are limits to its effectiveness. As these bossless firms grew, they faced increasing pressure to adopt a more hierarchical structure. While some companies (like Valve), laid off excess employees in order to maintain their bossless structure, some others (like Github) converted themselves into a traditional hierarchy. This made us think: what are the conditions under which it makes sense to be a hierarchical or a bossless organization? Should bossless firms adopt either form completely, or should they use other managerial levers? To address these questions, we built a simulation to model the core processes of these innovative firms in both hierarchical and bossless forms: evaluation of opportunities and allocation of employees to them. By doing so we were able to compare the effectiveness of these hypothetical firms (both bossless and hierarchical) under different conditions.

We found that the key factor that determines effectiveness is the balance between available human resources and available opportunities. When the number of available opportunities is much greater than the human resources under its control, bossless firms perform much better when compared with a hierarchy. A boss would be conservative, and biased in favor of certain projects, thus making the firm lose out on lucrative ones. However, as the organization grows, allowing employees to self-select themselves into projects leads to overstaffing (more resources than needed on good projects) and resource misallocation (employees choosing wrong projects). Here, a boss helps in reducing the winners' curse of good projects which might get overcrowded, by being conservative.
Of course, changing overnight from a bossless structure to a hierarchy is nearly impossible or highly risky, so bossless firms usually adopt various policy levers to limit the pathologies produced by self-selection. These include: a) allowing employees to leave previously selected projects freely; b) setting a minimum number of workers necessary to launch a project, c) setting up a minimum required profit threshold, d) requiring an approval from a superior to launch a self-initiated project, and e) finally introducing additional incentives for managers to boost their project evaluation and employee allocation effort. We simulated these policy levers with our model and found that even these levers have limits. Under certain conditions these levers become ineffective, or even present adverse effects.

Why should this matter for managers? Firstly, this choice of structure presents implications for both young and mature firms alike. Typically, young startups are seen as flat and innovative (good), and mature firms are seen as hierarchical and conservative (bad); however, we suggest that the choice of form is dictated by the balance between resources and opportunities. Overstaffed skunkworks that give employees a radical amount of autonomy may in fact be less beneficial for the firm than centralized units that explore new ideas or innovation. Secondly, as a firm grows or the industry matures, the low-hanging fruit of innovation are already taken, that is, lucrative projects are no longer abundantly available. At this point, it may no longer be feasible for the organization to stay bossless; it may be eventually compelled to adopt some form of hierarchy. Thus, we aim to inform managers on choosing the right organizational structure for their firms.