Every manager says they want a high-performing team. Fewer managers have a clear, evidence-based understanding of what high performance actually requires and what they need to do consistently to build it. The result is a great deal of effort spent on the wrong things: team away days that feel good but leave team dynamics unchanged; performance conversations that address output without touching behaviour; hiring decisions that optimise for individual brilliance rather than team contribution.
The research on what makes teams high-performing is robust and consistent. The conditions are well understood. What varies is whether leaders are willing to do the harder, less visible work of building those conditions.
The Foundations: What the Research Shows
Google's Project Aristotle studied hundreds of internal teams across the organisation over several years to identify what distinguished high performers from average performers. The finding was counterintuitive for a company that prized individual brilliance: team composition mattered far less than team dynamics. Specifically, psychological safety — the shared belief that it is safe to take interpersonal risks — was the single strongest predictor of team performance.
Amy Edmondson's three decades of research at Harvard Business School confirm and extend this. Teams with high psychological safety surface problems earlier, share information more freely, learn from mistakes more effectively, and consistently outperform teams where members self-censor. The mechanism is straightforward: when people feel safe to speak up, the information needed to perform well actually gets shared. When they don't, it doesn't.
Beyond psychological safety, the research identifies several other consistent conditions for high team performance. Clear goals and role clarity: every team member understands what the team is trying to achieve and what their specific contribution to it is. Genuine interdependence: the work actually requires people to collaborate, rather than just reporting to the same manager. Appropriate tools and resources: teams cannot perform without what they need to do the work. A supportive organisational context: the team's goals, structures, and rewards align rather than conflict.
What Leaders Actually Do
Understanding the conditions is not sufficient. The question for a manager is: what do I specifically need to do to build and maintain them?
Psychological safety starts with you. The research is consistent: psychological safety is built or eroded primarily through leadership behaviour. Specifically, it is built when leaders model vulnerability — acknowledging uncertainty, admitting mistakes openly, asking for input before expressing their own view, and treating challenge and disagreement as valuable contributions rather than threats. It is eroded when leaders respond to bad news defensively, punish candour, or signal that their view is the right one before discussion has happened. No amount of team workshops will build psychological safety if leadership behaviour is consistently unsafe.
Clarity is a leadership output, not a meeting outcome. Many teams lack goal and role clarity not because their leaders haven't discussed it, but because the discussion produced different conclusions in different people's minds. Clarity requires externalising and checking understanding. Ask each team member to articulate the team's priority goals and their own specific contribution to each. The gaps between what you said and what they heard will immediately reveal where you need to be clearer. This process needs to be repeated, particularly as priorities shift.
Performance conversations need to happen more frequently and be more specific. One of the most consistent findings in the management development literature is that managers delay or avoid performance conversations, either because they don't have the skills to hold them well or because they fear damaging the relationship. High-performing teams are built in part by managers who have short, specific, frequent feedback conversations that are depersonalised and behavioural. The longer you wait to address performance, the harder the conversation becomes and the less useful the feedback is.
Team decisions need to be made with the team. Research by Hackman (2002) found that one of the most common mistakes managers make is deciding alone when they should be deciding collaboratively. This matters for two reasons. First, collaborative decisions tend to be better decisions because they draw on more information. Second, people are more committed to implementing decisions they have had meaningful input into. The skill is knowing which decisions require full team involvement, which require consultation without delegation, and which you should make alone.
Handling Underperformance
The hardest test of any manager's commitment to high performance is how they handle sustained underperformance. The research is clear that allowing underperformance to continue without intervention damages team performance significantly — not just through the direct output impact but through the signal it sends to the rest of the team about standards and accountability.
Effective intervention starts with understanding the cause. There are three broad reasons why people underperform: they lack the capability to do what is required (skill gap), they lack the motivation to do it (will gap), or something in their environment is making it harder than it should be (situational gap). Each requires a different response. Providing coaching to someone whose problem is situational will not help. Setting performance targets for someone with a genuine skill gap is unfair.
Once the cause is clear, the response should be specific, time-bounded, and supportive. What specifically needs to change? By when? What support will you provide? What happens if things don't improve? These questions, asked directly and answered honestly, are the structure of an effective performance conversation. Avoiding them, however understandable the discomfort, means the situation will get worse.
Try This
Conduct a brief team effectiveness audit this week. Ask each member of your team, individually and confidentially, two questions: What is one thing we do particularly well as a team? What is one thing that most gets in the way of our performance? The answers will give you an immediate picture of where psychological safety stands and what the most important development priorities are. Share a synthesised version of the themes with the whole team and ask them to help design the response.
References
Edmondson, A.C. (1999) 'Psychological safety and learning behaviour in work teams', Administrative Science Quarterly, 44(2), pp. 350–383.
Edmondson, A.C. (2023) Right Kind of Wrong: The Science of Failing Well. London: Atria/One Signal Publishers.
Hackman, J.R. (2002) Leading Teams: Setting the Stage for Great Performances. Boston, MA: Harvard Business School Press.
Rozovsky, J. (2015) The five keys to a successful Google team. Google re:Work.
Wageman, R. et al. (2008) Senior Leadership Teams: What It Takes to Make Them Great. Boston, MA: Harvard Business School Press.