When It's Not One Person but the Whole Team: Recognising a Collective Dynamic

I return to it at greater length in Your Thoughts Don’t Have to Run Your Life.

A team doing badly always produces an explanation, and that explanation is almost always a person. This is not malice: it is the shortest route. An organisational cause requires analysis; a human cause is visible immediately, is easy to recount, and gives the impression that action is possible.

The problem is that it is wrong in a great many cases — and that it has a precise cost: the person designated leaves or moves, and six months later the same tension reappears, with somebody else.

Three signs that this is a dynamic, not a person

These three signs are observable and call for no interpretation.

The tension moves. Somebody leaves, the situation eases for two months, then the same tension reappears — at the same point in the process, on the same subject, with a different person. This is the most decisive sign, and it is only visible with hindsight. It reproduces across different pairs. If three different pairs of people clash over the same object — sign-off, prioritisation, access to a resource — it is not a matter of compatibility. It follows a calendar. Tension rises at quarter-end, during reporting week, before each committee. A rhythm is the signature of a constraint, not of a relationship.

These signs can be recorded, and they record far better in the exchanges than in memories. The density of messages by period, the channels used, the concentration of exchanges on certain members, the timings: that is what ScanMyJob puts into series, and the team conflict case shown as an example — seven participants, thirteen weeks — shows exactly this kind of reading: what share of the exchanges concentrates on two contributors, what share goes through the group channel, at what moments messages fall outside working hours.

The six configurations

1. Contradictory objectives between functions

What it is: two people are assessed on indicators that cannot be met together. Sales on signed volume, delivery on margin; production on lead time, quality on the rework rate; operations on availability, safety on shutdowns. The indicators: the disagreement is always the same, it comes back on every file, and both positions are defensible. Nobody is wrong, which is precisely the signature. What acts on it: nothing relational. Only a change of objectives or the addition of a shared indicator changes anything, and that is decided above the two people concerned. Bringing them together to "work on their relationship" asks them to resolve between themselves a contradiction the organisation created.

2. The single contested resource

What it is: a skill, a tool, a person, a slot that several people depend on and that cannot serve everyone. A sole expert, a cross-functional developer, a room, a machine, a budget. The indicators: the tension bears on access, not on substance. Requests come earlier and earlier, then more and more insistently, then by roundabout routes. The person who is the resource is often the first to be exhausted, and the last to be seen as concerned. What acts on it: an explicit allocation mechanism — a queue, announced prioritisation, arbitration by a third party — or an increase in the resource. Nothing else. In the absence of a rule, access is settled by relative force, and it produces exactly the behaviours that will afterwards be held against the people.

3. Overlapping remits

What it is: two functions whose boundaries have never been settled. Who answers the client, who signs off, who is informed, who decides in the event of disagreement. The indicators: they are formal before they are verbal. Sign-offs requested twice, documents in parallel versions, meetings called on the same subject by two people, and one recurring sentence: "I wasn't aware of that." What acts on it: a written clarification, with the overlapping areas named and an arbitrator designated for them. One hour of laying things out often settles what six months of tension has not. This is the most frequent configuration and the easiest to deal with — provided it is identified before it has migrated into the relationship, which takes a few weeks.

4. The absence of decision-making

What it is: decisions are not taken, or are taken and then reopened. The team then has to renegotiate permanently what ought to be settled. The indicators: the same subjects come back to meetings without conclusion; decisions circulate orally and contradict one another; people build up proof of what was said. That last point is very characteristic: when a team archives systematically, it is because it cannot rely on what is decided. What acts on it: a record of decisions, written, dated, with a name attached. It is the simplest remedy on the list and one of the most effective. It requires no difficult conversation.

5. Workload

What it is: the volume of work is lastingly greater than what the team can absorb. Every request then becomes an aggression, not out of malice but by arithmetic. The indicators: tension rises with the load and falls with it; requests are refused before they have even been read; messages go out beyond working hours; mutual help, which is the first indicator of a team doing well, is the first thing to disappear. What acts on it: the workload, and nothing else. Any relational intervention on an overloaded team is perceived — rightly — as displacing the problem. It is also the configuration in which the risk to people's health is highest: if your sleep, your appetite or your recovery have lastingly changed, an occupational health doctor can be seen without going through the employer and is bound by medical confidentiality.

6. Instability of composition

What it is: arrivals and departures follow one another, and the team never gets past the stage at which implicit rules stabilise. The frame that explains it: the stages described by Bruce Tuckman — forming, storming, norming, performing — are sequential, and a team whose composition changes goes back a step. A tension that appears after an arrival is therefore not necessarily the sign of a problem with the person who arrived: it is the storming stage, which is a normal passage and is got through. The indicators: the tension has a date of onset coinciding with a change of composition, it bears on ways of doing things rather than on content, and it eases by itself within a few months if the rules settle again. What acts on it: making the rules explicit instead of waiting for them to be passed on. A team that changes often cannot run on the implicit.

What team research adds

Two contributions are worth knowing, because they shift the way you look.

Psychological safety. Amy Edmondson's work showed that one factor predicts a team's performance better than the competence of its members: the possibility of saying there that you do not know, that you got it wrong, or that you disagree, without fearing you will pay for it. A team in which error cannot be voiced does not become a team without errors: it becomes a team in which errors are discovered late.

The marker is simple and can be checked: the last time somebody got something wrong, did they say so themselves, or was it discovered?

The cascade of dysfunctions. Patrick Lencioni describes a chain that repays reading backwards: inattention to collective results comes from avoidance of accountability, which comes from lack of commitment, which comes from fear of conflict, which comes from absence of trust. A team in which nobody ever contradicts anyone in a meeting is not a team at peace — very often it is a team in which disagreements are handled elsewhere.

Who can do what

A team member can do three things, and they are not negligible. Describe the configuration instead of designating a person — "requests from department X reach both of us with no priority set" rather than "he takes everything". Document the regularity, which is the only argument that moves a manager. And take the question forward as an organisational request, which is the only register in which it is actionable. A manager can act on five of the six configurations: remits, decision-making, resource allocation, explicit rules, and — partly — workload. The approach has to be prepared, and the first step is always the same: check whether the tension moves, reproduces, or follows a calendar. The level above is the only one able to deal with the first configuration — contradictory objectives between functions — and it is the one that wears people down longest, precisely because nobody at the level concerned can do anything about it.

To take stock of what the situation is costing you and of your way of approaching collective tensions, the tests on the platform are available, and the assistant can help you word a description of a configuration before taking it forward.

Three things this article does not do

It does not say that nobody is involved. Individual behaviours exist and they count. What it argues is that the three signs should be checked before concluding — because getting the cause wrong costs a departure for nothing. It does not describe anyone. The six configurations describe properties of an organisation, not characteristics of team members. That is what makes them treatable. It characterises nothing. Some situations belong to other circuits — employee representatives, occupational health doctor, HR, a lawyer. Assessing them belongs to a court, not to an article.