Rival CEOs resign on a Friday, boards rush to name a savior by Monday, and the stock sinks by Wednesday. The ritual persists even as work itself has changed. Knowledge-intensive organizations run on interdependence, not solitary brilliance. The hard question for 2026 is whether succession should mirror that reality. Instead of betting the firm on a single hero, can companies rotate the chair and share power without sowing confusion?
What the evidence actually shows
Two independent meta-analyses reach the same point. When leadership is shared across a team, performance improves. D'Innocenzo, Mathieu and Kukenberger reviewed 50 effect sizes from 3,198 teams and found a positive relationship between shared leadership and team performance, with stronger effects in real organizational settings than in classroom or lab studies (D'Innocenzo, Mathieu and Kukenberger, 2016). How we measure sharing matters. Studies that map influence as a network, such as leadership density or decentralization, show larger effects than simple averages of self-reported behaviors.
A second meta-analysis by Wang, Waldman and Zhang examined 42 independent samples and estimated an overall correlation of rho = .34 between shared leadership and team effectiveness (Wang, Waldman and Zhang, 2014). The type of influence matters here too. Sharing so-called new-genre leadership, such as transformational or charismatic behaviors, relates more strongly to effectiveness at rho = .34 than sharing traditional initiating-structure or consideration behaviors at rho = .18. The relationship also strengthens as work becomes more complex, which is precisely the context facing most knowledge-work teams.
Precision is important. These are team-level findings. They do not prove that a rotating CEO or a co-CEO arrangement will outperform a single chief in every firm. They do show that when influence is distributed and visible in the fabric of work, performance tends to be better, especially in complex tasks and in field settings that resemble day-to-day business. That is a useful anchor for designing succession that relies on a bench of leaders rather than a single anointed figure.
From the team room to the boardroom
At the very top, the evidence is thinner but more encouraging than conventional wisdom admits. A study summarized in Harvard Business Review analyzed 87 companies led by co-CEOs and concluded that the model is workable, though rare (Feigen, Jenkins and Warendh, 2022). Among roughly 2,200 companies in the S&P 1200 and Russell 1000 between 1996 and 2020, fewer than 100 employed co-chiefs during that period. The authors note real failures, pointing to Chipotle Mexican Grill, SAP, and Research In Motion/BlackBerry as examples of jointly led companies that performed poorly, especially under stress. Their central claim is not that co-CEOs outperform single bosses. It is that the arrangement can be more viable than most boards assume when designed with care.
Reconciling the two literatures leads to a practical stance. Teams benefit when leadership is a networked property, not a title. Co-CEOs can work, but the bar for clarity and coordination is high. That suggests boards should build succession systems in which multiple leaders practice taking the lead on real work, and in which the organization can switch who holds the formal authority without losing coherence.
How to design rotating leadership that actually works
-
Build leadership density, then decide who rotates. The network-based measures that show stronger performance effects capture how widely influence is shared inside a team (D'Innocenzo, Mathieu and Kukenberger, 2016). Start by spreading decision-making and ownership across roles. Rotation should surface that existing influence, not paper over a vacuum with job swaps.
-
Match the model to complexity. The performance benefits of shared leadership intensify as work becomes more complex (Wang, Waldman and Zhang, 2014). That argues for piloting rotating or shared leadership in knowledge-heavy units first, such as product, R&D, or platform operations, where interdependence is highest and the upside is largest.
-
Elevate new-genre leadership behaviors. Because sharing transformational behaviors relates more strongly to effectiveness than sharing traditional supervisory routines, invest in how leaders frame purpose, set direction, and coach peers, not just how they assign tasks (Wang, Waldman and Zhang, 2014). Rotations should be anchored in these higher-order behaviors.
-
Codify decision rights and external voice. The co-CEO record includes notable failures under stress, which is a cue about ambiguity, not a verdict on sharing power (Feigen, Jenkins and Warendh, 2022). When rotating leaders or pairing chiefs, write down who decides what, how tie-breaks occur, and who speaks to investors, regulators, and the press. Ambiguity that is tolerable inside a team can become costly at the boundary of the firm.
-
Use network-informed diagnostics. Because social network approaches better capture the performance link, measure who colleagues actually turn to for expertise, coordination, and escalation, then align rotations with those informal maps (D'Innocenzo, Mathieu and Kukenberger, 2016). This reduces the risk of elevating a formal title while sidelining the real node of influence.
-
Treat rotation as real succession practice, not theatre. Assign rotating leaders P&L authority, accountability for a cross-functional mission, and a visible review cadence. The aim is to test whether the organization performs when the formal head changes, not to run a tour of duty for résumés. Frame each rotation around outcomes the board already tracks.
The practical read for boards and CEOs
If your operating core is complex and interdependent, the odds favor shared leadership at the team level. Two meta-analyses show positive associations with effectiveness in real organizations, stronger under complexity, and stronger when leadership is understood as a network property (D'Innocenzo, Mathieu and Kukenberger, 2016; Wang, Waldman and Zhang, 2014). At the top, co-CEOs remain rare, and there are notable failures, but a study of 87 such companies argues that shared top leadership is more viable than its reputation suggests (Feigen, Jenkins and Warendh, 2022).
Taken together, the evidence points to a disciplined experiment. Build leadership density, use network diagnostics to pick candidates, codify decision rights, and rotate authority over real work. Report outcomes against the same metrics you would apply to a permanent appointee. You may still appoint a single chief. You will do so with a stronger bench, a cleaner handover, and fewer surprises. More to the point, you will have designed succession as an operating capability, not a once-a-decade cliff event.
Sources
- D'Innocenzo, Mathieu and Kukenberger (2016). A Meta-Analysis of Different Forms of Shared Leadership–Team Performance Relations. Synthesizes 50 effect sizes from 3,198 teams and finds a positive link between shared leadership and performance, stronger in field settings and when measured as a network. https://journals.sagepub.com/doi/10.1177/0149206314525205
- Wang, Waldman and Zhang (2014). A meta-analysis of shared leadership and team effectiveness. Estimates an overall correlation of rho = .34 between shared leadership and team effectiveness, stronger under complex work and for shared transformational leadership. https://pubmed.ncbi.nlm.nih.gov/24188392/
- Feigen, Jenkins and Warendh (2022). Is It Time to Consider Co-CEOs? Reviews 87 co-CEO companies, notes the rarity of the model and notable failures, and argues the arrangement is more viable than commonly assumed. https://hbr.org/2022/07/is-it-time-to-consider-co-ceos