Boards still treat succession as a cliff edge. A resignation lands, a shortlist appears, and the firm bets billions on a single name. The pattern repeats even as work grows more interdependent and complex. There is a better path hiding in plain sight: make leadership a rotating practice, not a one-off event.
What the evidence actually says about sharing the seat
Two independent meta-analyses converge on a simple claim. When leadership is shared across a team, performance tends to be higher. In a review of 50 effect sizes across 3,198 teams, D'Innocenzo, Mathieu and Kukenberger (2016) find a positive relationship between shared leadership and team performance, with stronger effects in real organizations than in classrooms or labs. Their study also reports that how you assess shared leadership matters. Social network measures of who actually influences whom, such as density and decentralization, show larger effects than simple averages of individual leadership ratings (D'Innocenzo, Mathieu and Kukenberger, 2016).
A separate synthesis by Wang, Waldman and Zhang (2014) reports an overall corrected correlation of 0.34 between shared leadership and team effectiveness across 42 independent samples. The effect is not uniform. Sharing so-called new-genre leadership behaviors associated with vision and transformation correlates more strongly with effectiveness than sharing traditional initiating-structure or consideration behaviors. Crucially for modern knowledge work, the positive association is stronger when the team's work is more complex (Wang, Waldman and Zhang, 2014).
These are team-level results. They do not prove that rotating a CEO will lift a firm's stock price. They do show that in environments where interdependence and complexity are high, distributing influence improves outcomes, particularly when you can see and tune the underlying influence network. Succession design can draw directly on that logic.
Co-CEOs are a living lab, not a magic wand
Skeptics will point to joint chiefs who sank under pressure. They are right to warn about stress tests. Chipotle Mexican Grill, SAP, and Research In Motion all struggled under co-CEOs, especially when conditions turned adverse (Feigen, Jenkins and Warendh, 2022). The arrangement is rare to begin with. Among roughly 2,200 companies in the S&P 1200 and Russell 1000 between 1996 and 2020, fewer than 100 used co-chief executives (Feigen, Jenkins and Warendh, 2022).
Rarity is not the same as impossibility. In their study of 87 companies with co-CEOs, Feigen, Jenkins and Warendh (2022) argue the model is more viable than most boards assume. The lesson for succession is not to copy a label. It is to use co-leadership as a rehearsal space where influence is shared on purpose, decision rights are explicit, and the organization practices continuity while leaders are still in the building. That is how you surface the failure modes early: fuzzy accountability, culture splits, or political triangulation. Shared leadership evidence suggests the payoff is likelier when the work is complex and when you can map and manage the real network of influence rather than rely on titles alone (Wang, Waldman and Zhang, 2014; D'Innocenzo, Mathieu and Kukenberger, 2016).
Design rotation to teach, not just test
If shared leadership correlates with stronger outcomes at the team level, and if co-CEOs can work with the right design, then the practical move is to build rotation into the operating model.
- Rotate decision rights, not just jobs. Give rising leaders time-limited ownership of real decisions with material stakes, paired with a peer who has reciprocal authority. This recreates the conditions under which shared leadership shows its strongest links to performance: influence spread across a network in complex work (Wang, Waldman and Zhang, 2014).
- Make the influence map visible. Use social network analysis to assess where people actually go for guidance, challenge, and approval. D'Innocenzo, Mathieu and Kukenberger (2016) report stronger performance links when shared leadership is measured as a network. That implies leaders should manage the network, not just the org chart.
- Stage co-leads with explicit seams. Co-ownership of a product line or region can be a low-drama way to test complementary styles. The co-CEO evidence shows the model is rare and not foolproof, with real failures, so clarity on who decides what and when is non-negotiable (Feigen, Jenkins and Warendh, 2022).
- Shorten the feedback loop. Shared leadership works best when people can see cause and effect. Tie rotations to crisp metrics and frequent reviews, so influence shifts based on outcomes rather than tenure or charisma. The meta-analytic pattern is correlation, not causation. That is a reason to measure and adjust in cycle, not a reason to avoid the practice (Wang, Waldman and Zhang, 2014).
The practical read for boards and CEOs
Succession that depends on a lone heir is fragile. The evidence base favors systems that practice power-sharing before a crisis. Meta-analyses show that distributing leadership within teams is associated with stronger performance, especially when the work is complex and when the real influence network is dense and not overly centralized (Wang, Waldman and Zhang, 2014; D'Innocenzo, Mathieu and Kukenberger, 2016). At the top, co-leadership is rare and carries risks, yet it is viable under the right conditions, which makes it a credible instrument for rehearsal rather than a last-ditch fix (Feigen, Jenkins and Warendh, 2022).
The implication is concrete. Treat succession as a muscle. Map the influence network, assign overlapping decision rights, and rotate leaders through real authority while pairing them with peers whose strengths differ. Use those rotations to debug the seams that have undone famous co-CEO experiments. By the time a vacancy appears, you will not be auditioning strangers. You will be choosing among people who have already done the work, together, in the daylight.
Sources
- D'Innocenzo, Mathieu and Kukenberger (2016). A Meta-Analysis of Different Forms of Shared Leadership–Team Performance Relations, Journal of Management. Finds a positive link between shared leadership and team performance, stronger with network measures and in field settings, supporting rotation as a system-level design. https://journals.sagepub.com/doi/10.1177/0149206314525205
- Wang, Waldman and Zhang (2014). A meta-analysis of shared leadership and team effectiveness, Journal of Applied Psychology. Reports an overall rho of .34 for shared leadership and stronger effects in complex work, reinforcing the case for distributing influence in knowledge-intensive contexts. https://pubmed.ncbi.nlm.nih.gov/24188392/
- Feigen, Jenkins and Warendh (2022). Is It Time to Consider Co-CEOs?, Harvard Business Review. Shows the arrangement is rare, notes notable failures, yet argues co-CEOs are more viable than many assume, offering a living lab for shared top leadership. https://hbr.org/2022/07/is-it-time-to-consider-co-ceos