Belbin team roles

Meredith Belbin developed the model at Henley in the 1970s, watching teams of managers compete in a business simulation over several years. His team could not predict which groups would win from intelligence scores alone. What predicted performance better was the mix of behaviours in the room. The findings were published in Management Teams: Why They Succeed or Fail in 1981, with eight roles; the ninth, Specialist, was added in later work.

One early result is still the most useful thing in the book. Teams assembled entirely from the sharpest analytical minds performed badly. They debated brilliantly, converged on nothing, and lost to ordinary teams. Belbin called this the Apollo syndrome, and most managers recognise it the moment they hear it.

The nine Belbin team roles

The three groups, and what each is for

The nine roles sort into three clusters. Reading the clusters is faster than memorising the roles, and it tells you where a team is thin.

• Thinking roles (Plant, Monitor Evaluator, Specialist) supply ideas, judgement and expertise. A team without them repeats last year's answer with more energy.
• Action roles (Shaper, Implementer, Completer Finisher) supply momentum, structure and finish. A team without them has a superb strategy and nothing shipped.
• People roles (Coordinator, Teamworker, Resource Investigator) supply direction, cohesion and outside contact. A team without them fragments, or quietly becomes an island.

Most dysfunctional teams are not short on talent. They are stacked in one cluster. Three Shapers will fight for the wheel. Three Monitor Evaluators will still be evaluating in the second quarter.

"Allowable weakness" is the part everyone skips

Belbin's sharpest idea is that each strength arrives attached to a specific cost, and the cost is the price of the strength rather than a defect to be trained away. The Plant who ignores detail is the same Plant who ignores the constraint everyone else accepted. Fix the first and you often lose the second.

A weakness is allowable when the team gets more from the strength than it loses from the cost. It stops being allowable when it goes unmanaged.

So the managerial move is not correction. It is compensation: put a Completer Finisher near the Plant, give the Shaper a Teamworker to repair the wake, and stop asking the Monitor Evaluator to be enthusiastic. Belbin also distinguishes an allowable weakness from a non-allowable one, where the behaviour has run past its usefulness. A Teamworker who avoids confrontation is normal; a Teamworker who suppresses a real disagreement until it becomes a resignation is not.

The Belbin test: what it does and does not measure

The instrument most people mean by "the Belbin test" is the Self-Perception Inventory, in which you distribute points across statements about how you behave at work. Belbin's own approach pairs it with Observer Assessments from colleagues, and this pairing matters more than the questionnaire itself. Self-perception is a hypothesis. Observer data is evidence.

• It measures behaviour in a work context, not personality type. Your roles can shift with a different team, a different job, or a different manager.
• You get preferred roles, manageable roles and least preferred roles. Most people have two or three they play naturally.
• Scoring is ipsative: your points are shared out across the roles, so a score describes your own ranking rather than comparing you to anyone else. Two people with the same top role are not equally strong at it.
• It is not a selection tool. Using Belbin scores to hire, promote or exclude someone is a misuse, and Belbin Associates say so directly.

How to use Belbin with a real team

1. Map the team before you map individuals. List the nine roles and mark which are covered, thinly covered or absent. The gaps are the finding.
2. Check the gap against the work in front of you. A team missing Completer Finisher is fine on exploratory work and dangerous three weeks before a launch. Role gaps only matter relative to the task.
3. Assign the missing contribution as a job, not an identity. Someone can run the Monitor Evaluator function for one decision without being one. Rotate it.
4. Use the language to depersonalise conflict. "We have two Shapers and no Coordinator on this" lands better than "you two argue too much", which is why the model helps most in feedback conversations.
5. Re-run it after the team changes. Roles adapt. People often step into a vacated role when the person who held it leaves.

One practical warning: run the mapping as a team activity, not a private diagnosis you deliver. The insight comes from people recognising themselves out loud, and that requires enough psychological safety for someone to admit which weakness is theirs.

Where the model breaks down

Belbin is a useful lens with real limits. Four worth stating plainly:

1. The questionnaire's psychometrics have been contested. Researchers including Adrian Furnham and colleagues published critiques of the Self-Perception Inventory's internal consistency in the early 1990s, and Belbin responded. Treat individual scores as a conversation starter, not a measurement.
2. Ipsative scoring cannot compare people. Because points are distributed within a person, you cannot say one person is "more of a Plant" than another. Team-level charts that stack individual scores side by side overstate what the data supports.
3. Labels calcify. The most common failure in practice is a team that has been Belbin-tested once and now uses the roles as excuses. "I'm a Plant, I don't do detail" is the model being used to avoid growth rather than to organise around it.
4. Nine roles rarely fit a small team. With five people you cannot cover nine roles, and you should not try. Cover the clusters, not the grid.

There is also a claim the model is often stretched into: that balanced teams reliably outperform unbalanced ones in every setting. Belbin's observations came from a specific simulation with management populations, and generalising them into a universal law is more than the evidence carries. What the model does reliably is give a team a shared vocabulary for a mix problem it can already feel.

Belbin alongside other models

The pairing that works best in practice is Belbin with Tuckman. Belbin tells you what the team is made of; Tuckman tells you where the team is in its life. A Shaper-heavy team in storming needs different handling from the same team in performing.

Seeing roles instead of reading about them

A questionnaire tells you what people believe about themselves. A task tells you what they do. Put a team into a short problem with a real deadline and the roles surface within minutes: someone starts organising before the brief is finished, someone spots the flaw everyone else skipped, someone quietly checks whether the person who stopped talking is alright. That is why experiential learning is the fastest route into this model. The debrief writes itself, because everyone in the room watched it happen. Run the inventory afterwards if you want a shared vocabulary, but run the task first.

Tour De Force runs this as live, experiential training for teams worldwide — online and in person. Talk to us, or play Gamified learning appsThe Weekly Challenge to see the method in ten minutes.