August 2, 2026 · Human Resources · 9 min read
Effective teams are not created by assembling a group of individually talented people and hoping chemistry appears. Team performance depends on how people coordinate, share information, manage interdependence, build trust, make decisions and learn together.
The strongest evidence does not point to one universal team formula. Different teams face different tasks, levels of uncertainty, interdependence and working arrangements. However, several recurring factors have good research support: trust and psychological safety, team cohesion, shared understanding, information sharing, reflection, clear goals and well-designed teamwork interventions.
A work group and a team are not identical.
Google’s Project Aristotle distinguished teams from groups partly through interdependence. A team depends on members coordinating their work to deliver a shared outcome. A group may report to the same manager or attend the same meetings while still doing largely independent work.
This distinction matters because many “team problems” are actually organisational-design problems. If people share no real outcome, have conflicting incentives and rarely depend on each other, team-building exercises will not create meaningful interdependence.
Source: Google re:Work, Understand team effectiveness
CIPD’s evidence review on high-performing teams draws together scientific research rather than relying on one company’s case study. Its main recommendations include attention to:
Source: CIPD, High-performing teams: An evidence review
Psychological safety refers to a shared sense that it is acceptable to take interpersonal risks such as asking questions, admitting mistakes, challenging an assumption or raising a concern.
CIPD’s evidence review reports a moderate-to-large positive relationship between psychological safety and team performance. It is closely related to trust but is not identical to it. Trust concerns what we expect other people to do; psychological safety concerns whether we feel able to speak or act without humiliation or punishment.
Source: CIPD, Trust and psychological safety: An evidence review
Psychological safety does not mean removing standards or avoiding accountability.
Managers can support it by:
A team cannot learn from problems that members are afraid to mention.
CIPD’s evidence review suggests team trust becomes particularly important under conditions such as:
Trust is strengthened by reliable behaviour over time. Clear commitments, follow-through, competence and fair treatment matter more than slogans about trust.
Teams need a shared understanding of:
When these are unclear, teams often compensate with more meetings, repeated approvals and duplicated work.
Google’s internal Project Aristotle studied team effectiveness at Google and identified five recurring team dynamics: psychological safety, dependability, structure and clarity, meaning, and impact.
This is a valuable organisational case study, but it should not be presented as a universal scientific ranking for every team in every sector. Google’s own guidance encourages organisations to examine the factors that matter in their own context.
Its practical value lies in the questions it prompts:
Teams perform poorly when knowledge is trapped with individuals or when important information reaches people too late.
Useful practices include:
The goal is not maximum communication. It is reliable access to the information required for coordination and judgement.
CIPD’s evidence review identifies team reflection as one element of effective teamwork. Teams need structured opportunities to examine how they are working, not only whether tasks were completed.
Useful review questions include:
A retrospective that produces no change becomes ritual rather than learning.
Cohesion can support trust, cooperation and shared identity. But a cohesive team can still make poor decisions if members become reluctant to challenge each other.
Managers therefore need both connection and constructive disagreement. Healthy teams should be able to challenge an idea without turning the disagreement into a personal attack.
Disagreement about a task, evidence or approach can be useful when it improves thinking. Relationship conflict, humiliation, bullying or unresolved hostility is different and can damage performance and wellbeing.
For a fuller distinction between ordinary disagreement, mediation and formal escalation, see MATSH’s Conflict Management at Work guide.
Smaller teams generally have fewer coordination links, while larger teams can contain more capacity and expertise. The appropriate size depends on the task, required skills, degree of interdependence, decision process and management structure.
Google’s Project Aristotle included teams ranging from three to fifty people, with a median of nine. That is descriptive of the teams in that study, not proof that nine is the universal optimum.
Instead of applying a fixed number, ask:
The forming, storming, norming and performing model remains a widely used way to discuss team development. It can help managers notice that team relationships and norms evolve.
It should not be treated as a scientifically precise sequence through which every team passes on a predictable schedule. Real teams may move backward, skip stages, change membership or operate differently under changing tasks.
Use the model as a conversation aid, not as a diagnosis that says every team must spend six or twelve months reaching a particular stage.
A meeting is useful when people need to make a decision together, solve a problem, coordinate interdependent work, learn from one another or build a shared understanding that would be difficult asynchronously.
Useful meeting discipline includes:
Individual manager conversations can cover priorities, blockers, feedback, development, working relationships and wellbeing. Gallup’s research on manager feedback also supports frequent meaningful conversations rather than relying only on infrequent formal reviews.
The frequency should reflect the employee and work. There is no universal rule that every person must have an identical weekly or fortnightly schedule.
Teams across the GCC, Africa and international organisations often include people with different languages, professional backgrounds and communication expectations.
Do not assume that every person from one country communicates or makes decisions in the same way. Instead, make team norms explicit:
Explicit norms reduce the need to interpret behaviour through cultural stereotypes.
Before organising another team-building event, ask team members to rate and discuss:
The discussion around different answers can be more useful than the score itself.
If a team is struggling, do not attempt to fix everything simultaneously.
A useful sequence is:
MATSH provides professional development in communication, management and team leadership. Training can help teams practise feedback, decision-making and communication, but effective teamwork also depends on job design, goals, incentives, information systems and manager behaviour.
There is no single universal factor for every team. Research supports multiple influences, including psychological safety, trust, clarity, cohesion, information sharing, reflection and goal setting. Google found psychological safety especially important in its own Project Aristotle study.
It is a shared sense that people can take appropriate interpersonal risks, such as asking questions, admitting mistakes or raising concerns, without being humiliated or punished for speaking up.
There is no universal ideal number. The right size depends on task complexity, interdependence, necessary expertise and coordination cost.
The Tuckman model can be a useful heuristic, but real teams do not always move through the stages in a fixed order or timeline.
Start with clear priorities, role clarity, decision rights and one or two coordination problems that create repeated friction. Visible improvements in those areas can create a foundation for deeper trust and team development.
We run all our courses as private programmes for organisations across the GCC and Africa.
Request In-House →