The pitch for the flat organization includes a quiet promise about succession: if there is no throne, there is no one to replace. Distribute authority across teams, let leadership rotate, remove the corner office, and the whole anxious drama of who takes over after the great leader leaves simply dissolves. No single point of failure, no succession crisis.
It is a good promise, and it is only half true. Removing the formal seat does not remove the influence, the relationships, and the accumulated judgment that clustered in it. Those migrate. And when they leave with a person, the organization can be caught more off guard than a hierarchy would be, because it told itself the person did not matter that much.
Distributed leadership is real, and so is its ceiling
The case for flattening is not fantasy. Meta-analyses find that shared leadership genuinely helps. Wang, Waldman, and Zhang, across 42 samples, found shared leadership related to team effectiveness at rho=.34, with a stronger effect on more complex work and an effect beyond what vertical leadership contributes. D'Innocenzo and colleagues, across 50 effect sizes and 3,198 teams, found a positive relationship that held more strongly in real organizations than in the lab. Authority spread across peers can lift performance, especially in knowledge work. That is a real finding, and it is why the flat model keeps attracting serious organizations.
But the same literature carries a caveat that matters for succession: the evidence is about teams, and extending it to a leaderless enterprise is an analogy, not a proven result. And a separate line of work is blunt about what happens at scale. Foss and Klein argue that firms which call themselves flat usually retain real hierarchy, just informal or renamed, and that in dynamic conditions the bossless ideal struggles. They point to Valve, whose flat structure produced what employees called creative inertia. The throne was not abolished. It was hidden.
The successor to the person nobody named
Here is the succession trap unique to flat organizations. In a hierarchy, when a powerful leader leaves, everyone knows a hole has opened and a replacement is needed. In a flat organization that has quietly concentrated influence in an unofficial figure, a founder who insists they are just another node, a veteran everyone actually defers to, that person's departure leaves a hole nobody is allowed to name, because the official story is that no such role existed.
Ricardo Semler's companies leaned hard into distributed authority, factory workers setting their own hours, teams able to veto a manager, engineers without degrees running plants on the strength of trust rather than title. That works, but it does not mean influence is evenly spread. Someone still carries disproportionate judgment and relationships. The honest version of a flat organization admits this and plans for the transfer of that informal authority, rather than pretending it does not exist.
Succeeding a system, not a seat
Succession in a flat organization is therefore a different task, not an absent one. What has to be handed over is not a title but a set of things a hierarchy would have bundled into a role: the relationships, the tacit judgment, the convening power that let one person align many others without an order. Those transfer badly by accident and reasonably well by design.
The shared-leadership evidence offers a genuine advantage here, if the organization uses it deliberately rather than as a slogan. Nicolaides and colleagues, in a meta-analysis, found that shared leadership lifts team performance partly through proximal emergent states such as trust and consensus. Those states are the connective tissue that lets a group absorb the loss of any one member without collapsing, because leadership was never sitting entirely in one place to begin with. A flat organization that has actually distributed authority, not just flattened the chart, has spread the risk of any single departure. That is the real succession dividend of flatness, and it is available only to the organizations that did the distribution for real.
The design questions are specific. Who actually holds the informal authority, regardless of what the chart says? What would break if they left, and who could absorb it? Is the organization developing the next people who could carry that convening weight, or is it relying on the myth of flatness to avoid the question? A flat organization that answers these is genuinely more resilient than a hierarchy. One that uses flatness as an excuse not to ask them is simply a hierarchy that has hidden its own single point of failure, and will find it the hard way.
Sources
- Wang, Waldman & Zhang (2014). A meta-analysis of shared leadership and team effectiveness. Across 42 samples, shared leadership related to team effectiveness at rho=.34, with stronger effects on more complex work and incremental effect beyond vertical leadership. https://pubmed.ncbi.nlm.nih.gov/24188392/
- D'Innocenzo, Mathieu & Kukenberger (2016). A Meta-Analysis of Shared Leadership-Team Performance Relations. Across 50 effect sizes and 3,198 teams, shared leadership related positively to performance, more so in real organizations than labs. https://journals.sagepub.com/doi/10.1177/0149206314525205
- Nicolaides et al. (2014). The shared leadership of teams: a meta-analysis. Shared leadership relates positively to team performance, partly through proximal emergent states such as trust and consensus. https://doi.org/10.1016/j.leaqua.2014.06.006
- Foss & Klein (2022). Who's the boss? The persistence of entrepreneurial hierarchy in flat organizations. Firms that call themselves flat usually keep real hierarchy, just informal or renamed; Valve's flat structure produced 'creative inertia.' https://link.springer.com/article/10.1007/s41469-022-00132-8