A leader is told to develop a successor and, if they take it seriously, immediately creates a problem for themselves. The deputy strong enough to take over one day is strong enough to be a rival today. They build their own following, form their own views, and start to look, to everyone watching, like the future. The leader who wanted to be responsible about succession has installed a plausible replacement one office away, and the organization can feel the ground shift.
The usual responses are both bad. One is to develop the deputy and then feel threatened, undermining them just when they are ready, which destroys the succession and often the relationship. The other is to avoid the problem by never developing anyone strong, which leaves the organization with no successor at all. Both are common. Both are failures.
Shared command is not equal command
One tempting resolution is to split the top, formalize the deputy as a co-leader, share the power so neither is subordinate. The evidence on this is more cautionary than the appeal suggests. Krause, Priem, and Love studied 71 co-CEO pairs and found a revealing shape: a larger power gap between the two was associated with better firm performance, up to a point, after which the gap became too wide and performance suffered. The relationship was an inverted U. Their reading favored unity of command over fully equal shared command, even inside a co-CEO structure.
The lesson for the second-in-command problem is precise. A deliberate, moderate power gap is not a flaw to be apologized for, it is what makes the arrangement work. The deputy needs real authority to grow and to be tested, but a structure that pretends the two are perfectly equal invites the paralysis and conflict the study warns about. Someone has to be, clearly, in charge, while the other is unmistakably being prepared.
Development requires safe disagreement
The deeper fix is cultural, not structural. A deputy is developed by being allowed to disagree, to make consequential decisions, and to be wrong in front of others without being punished for it. That only happens where challenging the boss is normal rather than dangerous.
Ricardo Semler's companies built that into the machinery. As he tells it, everyone rated their boss anonymously twice a year, and a manager who scored badly twice in a row lost the role, himself included. The team, not the boss's boss, could veto a manager. The first year was uncomfortable, but managers who used to shout started to ask. The point for succession is that in a place where a subordinate can safely challenge and even remove a bad boss, a deputy can grow into strength without it reading as insubordination or threat. The challenge is the system working, not a coup.
Threat is a signal about the leader
The leader who feels most threatened by a capable deputy is usually the one whose identity is fused with the position. If being the boss is who you are, then anyone who could be the boss is an existential rival. If being the boss is a job you are doing well and intend to hand on, a strong deputy is the best evidence you are doing it right.
It helps to name what the deputy actually needs in order to grow, because vague development is what breeds the rivalry in the first place. A deputy given real problems to own, with visible authority and clear accountability for the results, has a legitimate arena in which to build a reputation. A deputy given a title and no real mandate has only one way to demonstrate strength: by maneuvering against the boss for the authority they were denied. Much of what leaders read as a deputy turning into a rival is a deputy starved of real work looking for somewhere to prove themselves. Give them the work, and the rivalry usually has nowhere to grow.
The second-in-command problem, in the end, is less about the deputy than about the leader's willingness to be replaced. A leader secure enough to build someone strong, give them a real but subordinate mandate, and welcome their disagreement is running a succession. A leader who needs the deputy to stay safely weak is protecting an ego and starving the organization of a future. The rival you fear is often just the successor you have not made peace with.
Sources
- Krause, Priem & Love (2015). Who's in charge here? Co-CEOs, power gaps, and firm performance. In 71 co-CEO pairs, a larger power gap helped performance up to a point, then hurt it (inverted U), favoring unity of command over fully equal shared command. https://doi.org/10.1002/smj.2325
- Arena, Ferris & Unlu (2011). It Takes Two: The Incidence and Effectiveness of Co-CEOs. Early systematic study of shared top command in public firms, often tied to mergers, family firms, and transitions. https://doi.org/10.1111/j.1540-6288.2011.00305.x
- Frazier, Fainshmidt, Klinger, Pezeshkan & Vracheva (2017). Psychological Safety: A Meta-Analytic Review. Across 136 samples and 22,000+ individuals, psychological safety correlated strongly with learning behavior (rho=.62) and with positive leader relations (rho=.44). https://doi.org/10.1111/peps.12183