Executives arguing about where people work or how much autonomy teams should have are often fighting the wrong thing. The core variable is not office days or rituals. It is who gets to decide, on what, and when that authority shifts. Call an organization flat and power does not disappear. It goes informal, which is slower and harder to steer.
Why flatness tempts, and where it breaks
Self-managing organizations decentralize authority in formal and systematic ways, eliminating the manager–subordinate relationship as the default unit of control (Lee and Edmondson, 2017). The theory behind this design is clear. Managerial hierarchy is strong at reliable execution of known tasks in stable environments. It tends to falter when work is complex, nonroutine, and crosses boundaries, where status differences and rigidity can suppress learning and development (Lee and Edmondson, 2017). That promise draws leaders who want initiative and speed without the bureaucracy.
Yet structure is a choice under conditions. Flat or bossless models can function in stable contexts, but under dynamic and time-sensitive conditions they often struggle with innovation at scale and with longevity (Foss and Klein, 2022). Many firms that market themselves as flat keep real bosses. Titles soften, power clusters informally, and employees learn the unwritten map instead of seeing it (Foss and Klein, 2022). The cost is ambiguity. Decisions take longer, accountability blurs, and escalation becomes guesswork.
The early holacracy debate showed how quickly ideology can crowd out design. In 2016 Zappos was the most visible company to implement holacracy fully, and the debate around such experiments polarized into celebration or derision rather than a sober read of where the model actually helps (Bernstein, Bunch, Canner and Lee, 2016). What matters is not the label but the clarity of decision rights.
When hidden hierarchies slow you down
Valve’s experience illustrates what informal power looks like in practice. The company’s flat structure produced what insiders called creative inertia. By 2013 employees described its creative state as “The Wilderness,” a stall that is hard to square with the idea that no bosses means more initiative (Foss and Klein, 2022). The firm preserved flatness to the point of not building dedicated customer support and marketing departments. In a market that rewards fast customer response and coordinated launches, those are not cosmetic omissions. They are structural absences that slow learning and reduce the odds that a good idea ships on time (Foss and Klein, 2022). When Valve later produced Half-Life: Alyx, management intervention was required despite the bossless model, a signal that in moments of high coupling someone must hold the pen on trade-offs (Foss and Klein, 2022).
This pattern is not a paradox. It is the predictable result of replacing formal authority with informal authority. Decisions still get made, only through status, tenure, or coalition, which tend to be less transparent and often slower. Foss and Klein argue that near-bossless organizations are rare and, on average, do not outperform traditional firms. They also stress that managers matter in concrete functions: designing work, organizing resources, encouraging teams, and enforcing agreed rules (strategy+business, 2022). Their distinction between two kinds of authority is useful here. Authority Mark I covers the hard rights to hire, fire, and instruct. Authority Mark II covers leading, structuring, and building consensus. You can distribute more of Mark II and still reserve Mark I for defined moments. That is not a betrayal of autonomy. It is an operating model.
Zappos and the price of unclear power
Zappos made these trade-offs in public. In March 2015 the CEO gave employees a choice to fully adopt holacracy or take a buyout. Eighteen percent accepted the offer, a shock to any organization’s continuity (Quartz, 2020). By March 2017 the company replaced holacracy with an internal market in which teams ran as small businesses with their own P&Ls and transacted with one another at market prices, a move explicitly motivated by business performance challenges and criticism that holacracy was too inward-focused (Quartz, 2020). Over time Zappos brought managers back and stripped out the system’s more bureaucratic meetings, keeping mainly a circular chart as a remnant (Quartz, 2020). The lesson is not that self-management is folly. It is that structure must match the work and that decision rights must be legible. Markets inside firms can price some choices, but someone still sets the rules of the game and polices the edges.
The broader literature supports a contingent view. There is no universally superior structure. Bossless designs are a tool, not a creed, and formal hierarchy retains advantages when coordination must be tight and time horizons are short (strategy+business, 2022). That sits comfortably with the original case for self-management as a way to unlock cross-boundary problem solving and dampen status barriers, not as a universal substitute for managerial authority (Lee and Edmondson, 2017).
The practical read
Leaders do not need another label. They need a decision map. Start by specifying which decisions rest with teams by default and which require explicit Mark I authority. Publish who can stop work, commit funds above a threshold, reset priorities mid-sprint, or accept regulatory risk. Those are the choke points that create hidden bosses if you do not name them. Use role-based authority for most choices and person-based authority only for defined exceptions, a pattern that keeps initiative high while preserving a clear path to resolve conflict. Foss and Klein’s split between Mark I and Mark II gives you the vocabulary to do this without romanticism (strategy+business, 2022).
Then design for the weaknesses each structure introduces. If you reduce managerial layers, add forums that cut across functions so complex problems can be solved without waiting for a rescuer. If you keep more hierarchy for reliability, guard against the status effects that Lee and Edmondson warn can smother development and learning by making escalation psychologically safe and by rotating who leads cross-functional work (Lee and Edmondson, 2017). Finally, be honest about reversibility. Zappos showed that even at scale you can pivot from a purist model to a hybrid with clearer power, but the transition tax is real and paid in attrition and time (Quartz, 2020).
The test for any redesign is simple to state and hard to fake. In a real dispute, can a reasonable employee predict who decides, on what basis, and how quickly? If not, you do not have a flat organization. You have an ambiguous one. And ambiguity, not hierarchy, is the boss that keeps you from shipping.
Sources
- Lee and Edmondson (2017). Self-managing organizations: Exploring the limits of less-hierarchical organizing. Defines SMOs and explains when hierarchy helps or hinders, grounding the case for formal clarity in decision rights. https://www.sciencedirect.com/science/article/abs/pii/S0191308517300059
- Foss and Klein (2022). Who’s the boss? The persistence of entrepreneurial hierarchy in flat organizations. Shows how flat firms retain informal hierarchy and struggle under dynamic conditions, illustrated by Valve’s creative inertia and missing functions. https://link.springer.com/article/10.1007/s41469-022-00132-8
- strategy+business (2022). Can bossless management work? Summarizes Foss and Klein’s view that near-bossless firms are few and not superior on average, and clarifies managers’ functional roles and the Mark I vs Mark II authority distinction. https://www.strategy-business.com/article/Can-bossless-management-work
- Bernstein, Bunch, Canner and Lee (2016). Beyond the Holacracy Hype. Notes the polarized debate and identifies Zappos in 2016 as the largest full adopter, framing the context for later shifts. https://hbr.org/2016/07/beyond-the-holacracy-hype
- Quartz (2020). Zappos has quietly backed away from holacracy. Documents the 2015 ultimatum with 18% buyouts and the 2017 shift to an internal market, plus the later return of managers. https://qz.com/work/1776841/zappos-has-quietly-backed-away-from-holacracy