Belbin team roles: the nine roles and how to use them

Meredith Belbin's research began at Henley Management College in the 1970s with a genuinely good experimental design. Teams of managers competed in business simulations, and Belbin's group tried to predict which teams would win. Their first hypothesis — that the team with the highest collective mental ability would win — was wrong in an interesting way. Teams stacked with the sharpest individuals routinely underperformed. Belbin called them Apollo teams, and they lost because they spent their time in clever, unresolved argument and nobody organised anything.

What predicted success was the spread of behaviours in the team. Management Teams: Why They Succeed or Fail (1981) set out eight roles; the Specialist was added later, giving the nine used today.

The nine roles

Belbin groups the roles by orientation. The allowable weakness is the part that gets left out of most summaries, and it is the part that makes the model honest.

Thinking roles

People roles

Action roles

Three distinctions matter and are commonly lost:

1. A role is behaviour, not personality. Belbin measures how someone tends to behave in a team, which is influenced by personality but also by experience, the role's demands and who else is in the room.
2. People have preferred, manageable and least-preferred roles. Nobody is one role. You typically have two or three you naturally occupy, some you can do when needed, and one or two you should avoid.
3. Nine roles does not mean nine people. A team of five can cover the ground. What matters is that no critical role is entirely absent — most commonly the Completer Finisher in a team full of starters, or the Implementer in a team of ideas people.

Applied in a UK workplace

A local authority sets up a five-person project team to procure and implement a new booking system for its leisure centres.

The manager assembling it notices the obvious version of the team is three Shapers and two Specialists: everybody senior, everybody driven, two experts. This is Belbin's Apollo team, and the predictable failure mode is a fortnight of forceful argument about scope with no plan and no follow-through.

Instead:

• The Co-ordinator is the project lead — not the most forceful candidate, but the one who runs a meeting where everybody contributes and decisions get recorded.
• One Shaper provides drive, and the lead accepts the allowable weakness: this person will bruise someone during the procurement deadline, and that is managed rather than pretended away.
• The Monitor Evaluator is a finance colleague who is unpopular in workshops for being relentlessly sceptical, and who is exactly the person needed when three suppliers all claim their system integrates with everything.
• The Implementer is a centre supervisor who will turn the chosen system into a workable shift-level process.
• The Completer Finisher is the administrator who will find the twelve unmigrated membership records everyone else missed.

The interesting decision is the deliberate absence of a Plant. This is a procurement of existing systems, not an invention; adding a strong idea-generator would produce attractive proposals for building something bespoke that nobody has funded.

Two practical cautions from real use. First, do not announce roles as labels — "you're the Teamworker" becomes a licence ("I avoid confrontation, that's my role") and a slight. Talk about contributions the team needs and who is best placed to provide them. Second, the useful conversation is about gaps, not people: "nobody here is going to chase the detail — who is picking that up?"

Criticisms and limitations

• The Self Perception Inventory has real psychometric problems. Furnham, Steele and Pendleton's 1993 critique in the Journal of Occupational and Organizational Psychology questioned its factor structure and internal reliability; Belbin responded, and the debate has never fully resolved. Treat scores as a prompt for discussion, not a measurement of a person.
• It is self-report. Belbin's own answer — the observer assessment, where colleagues rate you — is better, and is the part organisations usually skip because it takes more effort.
• The original research base was narrow. Management simulations at Henley, with a demographically narrow sample by modern standards, generalised to all teamwork.
• Roles are not fixed, so the diagnosis has a short shelf life. People behave differently in a different team, under a different manager, or on work they care about more.
• It can become determinism. The model is most often misused to explain why someone cannot do something, which is the opposite of its purpose.
• It says nothing about competence or authority. A perfectly balanced team with the wrong skills, no budget or a goal that conflicts with the organisation's will still fail. Belbin addresses composition; Tuckman addresses development; neither addresses whether the task is achievable.

How to evidence this in your ILM portfolio

Belbin is a justification tool in a portfolio: it explains why you built or deployed a team the way you did.

Level 3 — Unit 315 Developing a high performing team and Unit 306 Developing project management skills. Evidence that works:

• A written analysis of your team's composition, the roles that are strong, and the gap you identified — with the action you took.
• Products of work: how you allocated tasks on a project, a revised team structure, a role-based pairing on a piece of work.
• A reflective account on a delegation decision made on behalf of team balance rather than availability, and how it turned out.
• Unit 305 Effective recruitment if you used role thinking in a real recruitment — being careful that selection decisions rest on the job requirements and competence, not on a questionnaire result.

Level 5 — Unit 503 Managing people for high-performing teams and Unit 509 Project management for project teams. At Level 5 the evaluation is the requirement: state the inventory's psychometric criticisms, say why you still found the framework useful, and be specific about where it misled you. That is a stronger answer than uncritical enthusiasm.

One compliance point worth noting: if you use any profiling tool on real colleagues, the results are personal data. Keep them out of your portfolio in identifiable form — anonymise, aggregate, or describe the composition without naming the person and their score.

Where to go next

• Tuckman's stages of group development — how a team matures once composed.
• Delegation: a manager's guide — allocating work to strengths in practice.
• Adair's action-centred leadership — the task, team and individual balance.
• Ready to qualify? Compare ILM Level 3 and ILM Level 5.