People often call me with a request phrased like this: « we should do something about team cohesion ». Almost always, when I dig, there is a specific conflict nobody dares name, or a decision taken two years ago that never landed. A request for cohesion is very often the polite form of a request for truth. So my first question is never « which activity? », but « what is not being said here? »
What we call « cohesion » — and what it actually is
Team cohesion is the force that holds a group together and moves it in one direction. It rests on three distinct components: shared commitment to the task, the quality of the bond between people, and pride in belonging to the group. It is not a pleasant atmosphere: it is a collective's capacity to hold together precisely when things get hard.
There are two very different things going by the same name.
The first is surface cohesion. Everyone gets on. Meetings are pleasant. Nobody raises their voice. It is comfortable — and it is often the sign that something is being avoided. A team with no disagreement is not a close team: it is a team where disagreement carries too high a price.
The second is real cohesion. It is recognisable by one simple test: what the group can go through without falling apart. A hard trade-off. A costly mistake. A fundamental disagreement between two key people. Genuinely cohesive teams are not the ones that avoid those moments — they are the ones that come out of them stronger.
« A team where nobody ever contradicts anybody is not close-knit. It is careful. That is not the same thing, and it does not produce the same results. »
| Surface cohesion | Real cohesion |
|---|---|
| Difficult subjects get avoided | They are put on the table early |
| Disagreements are settled in corridors | Disagreements are handled in the room |
| The relationship is protected by silence | The relationship is protected by candour |
| Good atmosphere, decisions drag | Demanding exchanges, decisions hold |
| It collapses at the first crisis | It strengthens in a crisis |
| It depends on personal affinity | It depends on the frame the group has given itself |
← Swipe the table to see all columns
What the research says — and what it doesn't
Cohesion is one of the most studied subjects in organisational psychology, and two findings are worth knowing before investing in anything.
Cohesion is not one thing. The meta-analysis by Beal and colleagues, published in 2003 in the Journal of Applied Psychology, shows that its three components — commitment to the task, interpersonal attraction and group pride — relate to performance independently of each other. In other words: improving the atmosphere without touching commitment to the shared work delivers a third of the possible effect. Which is exactly what a stand-alone cohesion off-site does.
The link is stronger the more people genuinely depend on each other. The same meta-analysis shows the cohesion-performance relationship strengthens markedly as workflow becomes intensely interdependent. A very concrete conclusion: before working on cohesion, check that the team has real reasons to work together. Many « cohesion problems » are in fact organisational design problems — people brought together whose work never actually intersects.
And the number one factor is not the one you'd expect. Google's Project Aristotle analysed more than 180 teams, ran over 200 interviews and examined 250 team attributes. The factor that most sharply separated the best teams from the rest was neither talent, nor seniority, nor size: it was psychological safety — the shared belief that you can take an interpersonal risk without being humiliated.
One honest caveat about this data: the correlation runs both ways. Cohesive teams perform better, but teams that succeed also become more cohesive. Which means a shared win often does more for cohesion than a day of activities — and a collective that achieves nothing will not be re-bonded over a barbecue.
« A request for cohesion is very often the polite form of a request for truth. »