Belbin Team Roles, reviewed

TL;DR

Belbin Team Roles, Reviewed: The Model That Argued Back

  • Born from nine years of real observation at Henley - a stronger origin than most instruments in this series.
  • The Apollo finding endures: teams of similar brilliance underperform balanced teams.
  • The psychometric debate was real and public; the publisher now calls its own test non-psychometric.
  • The balance insight survives; the nine fixed labels are where the model shows its age.

Most instruments in this series were born in a consulting room or a marketing meeting. Belbin's team roles were born in a business school laboratory, from one of the longer field observations management research has produced - which is why this review reads differently from the others, and why the model earned the right to a proper argument.

What are the nine Belbin Team Roles?

Plant, Resource Investigator, Co-ordinator, Shaper, Monitor Evaluator, Teamworker, Implementer, Completer Finisher and Specialist. Meredith Belbin's claim, from nine years of observing management teams at Henley, was that balanced teams - covering the thinking, people and action roles between them - outperform teams of individually brilliant but similar people. His famous Apollo experiment found teams assembled entirely from the sharpest analytical minds performed poorly, a result that still unsettles hiring instincts.

Where it came from

Through the 1970s, Meredith Belbin and his research team watched executives play competitive management games at what was then the Administrative Staff College at Henley, each participant profiled with a battery of psychometric tests beforehand. The programme's famous move was the Apollo experiment: assemble a team entirely from the highest-scoring analytical minds and watch it win. It lost, repeatedly - too much debate, too little execution, nobody willing to do the unglamorous work. Balanced teams of less individually stellar members beat the clever ones.

That result deserves its fame. It anticipated by decades the modern finding that how teams work together predicts more than who is on them, and it planted a permanently useful suspicion of the hire-the-cleverest instinct. A team role, in Belbin's definition, is "a cluster of behavioural attributes effective in facilitating team progress" - behaviour in context, not a fixed property of the person.

The argument in the journals

Is the Belbin test scientifically valid?

The record is genuinely contested. Furnham, Steele and Pendleton's 1993 assessment found poor psychometric properties in the Self-Perception Inventory, triggering a published exchange with Belbin himself. Later work was kinder: Swailes and McIntyre-Bhatty argued standard reliability statistics suit the inventory's forced-choice format badly, and a 2007 Journal of Management Studies review found qualified support for the model. Belbin's company now sidesteps the fight, stating the SPI 'does not have psychometric properties' because it measures behaviour, not personality - an unusual position for a widely sold assessment.

The 1993 exchange in the Journal of Occupational and Organizational Psychology - critique, reply, response to the reply - is worth respecting as a rarity: an instrument author engaging his critics in the open literature rather than in a brochure. Belbin's substantive defences have some force. The company's 2014 technical review reports observer-agreement data across thousands of observations and reliability figures in the .67-.84 range for the revised inventory, and the Swailes and McIntyre-Bhatty point about forced-choice formats distorting conventional reliability statistics is a legitimate technical argument, not mere spin.

The current positioning, though, invites a raised eyebrow. Belbin's FAQ states that reliability and validity "pertain to psychometric tests, and the Belbin SPI does not have psychometric properties" - while the same page cites peer-reviewed studies showing good validity. An instrument cannot straightforwardly claim validation's protection while declining its obligations. The generous reading is that the observer-assessment system - colleagues rating actual behaviour, corroborating or contradicting self-perception - is Belbin's real answer to the psychometric critique, and it is a better answer than most instruments in this series can offer.

What aged, and what did not

Is Belbin still relevant for team building?

Its core finding aged well: teams need complementary contributions, not clones of one ideal profile - a conclusion modern team research broadly supports from other directions. Its machinery has aged less well: nine fixed role labels compress the variety of real people, self-perception drives the scores unless observer data is added, and role language can harden into casting - the Plant excused from finishing, the Completer never asked for ideas. Use the insight about balance; treat the labels as conversation starters rather than job descriptions.

The casting problem is the one practitioners see most. Role vocabulary is sticky: within a season, the labels stop describing contributions and start allocating them. The Plant is forgiven every missed deadline; the Completer Finisher is never invited to the ideation day; the Shaper's abrasiveness becomes a job description. A model built to celebrate complementarity becomes a machine for excusing people from growth - the precise opposite of what Belbin observed at Henley, where roles were patterns in behaviour, revisable by evidence.

Keeping the finding, updating the instrument

Belbin's legacy finding - that teams succeed on complementary difference, not accumulated sameness - is one Sariio MAPS is built to serve, with the machinery updated. Instead of nine role labels, 60 work preference pairs on continuous scales; instead of a one-off inventory, a ten-minute survey retaken at least twice a year; instead of casting, a team map that shows where preferences cluster and where the gaps sit - who opens work and who closes it, where the ideas energy and the finishing energy actually sit in the room. The AI reads balance the way Belbin read his Henley teams, but per team, per quarter, with no fixed labels to harden into typecasting - and no verdict on any person. The related question of what any assessment can promise a team is picked up in personality tests at work.

Map one team's balance without the labels - the individual survey is free.

Sources