Why changing behaviour is hard

The usual reading is that the actions were not followed through. Sometimes that is true. More often they were followed through and the frustration came back anyway, which is a different problem and needs a different explanation.

Most improvement work follows the same sequence: gather the data, produce the insight, agree the actions. It is a comfortable sequence because every step has an owner and a date. It also has a poor record, because it treats behaviour as an information problem. Teams generally know which of their working relationships are not working; knowing was never what was missing.

A better description, and where it stops

If knowing is not what is missing, the next thing most people reach for is a better description of the problem. The most useful of those descriptions is the causal loop, which sets out what connects to what.

Something is designed centrally and pushed out to the business. In places it does not quite fit, and rather than say so, people go quiet and work around it. No objection reaches the centre, so the absence of objection is read as agreement, and the next thing built rests on the same assumptions. It fits even less, and even fewer people say anything.

That is an accurate account of a circle, and naming it does real work, because it moves the problem from a person to a pattern. What it cannot explain is why anyone continues to run it. A description of a habit is not an explanation of one.

The logic that keeps a pattern running

The obvious explanation is that the people inside the loop cannot see it. That is occasionally true and mostly not. People keep running these patterns knowingly, for reasons that are entirely rational from where they stand.

Someone holds information back because sharing it early has been used against them before. A team routes everything through its manager because the last time it acted independently it was overruled publicly. A difficult conversation stays unhad because the working relationship feels too fragile to risk it. Two departments compete for the same budget, so cooperation carries a quiet cost that neither will name in a meeting.

Each of these is a local optimum. The behaviour that damages the organisation is, for the person choosing it, the sensible option given what happened last time. That is why naming the pattern rarely ends it, and why fixing the symptom does not hold: the conditions that made the behaviour reasonable are still in place.

What the data reaches, and where it stops

The pattern itself is visible. Every Cophi survey collects free text alongside directional scores, and read together across a whole organisation those two show which patterns are running, in which teams, and how strongly. A team is not being told about organisations in general. It is being shown itself.

Two things follow. A team sees how it acts within the system, often for the first time. And because every team is measured against the same set, the patterns can be counted: when the same one appears strongly across forty teams, the problem does not sit inside those forty teams, it sits in something the centre is doing. That distinction between a local habit and a system-level one is not available from any single team's view.

Where it stops is precise. The data can name the pattern and show where it runs. It cannot say why that pattern pays for the people inside it. The overrule that is still remembered, the credit in a relationship too thin to spend, the rivalry nobody will name: none of it is recorded anywhere an instrument can reach. It only becomes available when it is said out loud.

The question that changes the conversation

If the reason has to come from the people inside the pattern, the question put to them decides what comes back. The usual question is who is causing this problem, and it reliably produces defence rather than description. The useful question is what pattern the group is collectively creating and sustaining.

That reframing is often heard as an invitation to be gentle. It is the opposite. Its purpose is to make honesty possible: if the pattern belongs to nobody in particular, people can afford to describe their own part in it accurately. Take the blame away and you can finally afford the truth. A conversation that stays comfortable by staying vague has missed the point entirely.

Why the conversation is difficult

Having the better question does not mean the conversation takes place. The obvious explanation for that is that people do not want to have it. In our experience that is rarely true. What is true is that the conversation requires a kind of attention organisational life is structured to prevent.

It needs to be slow, and calendars are built in half-hour blocks. It needs to stay open, and most meetings are judged by whether they closed with owners and dates. It needs people to describe their own part accurately, while most are present as representatives of a function, where doing so reads as conceding ground. And it needs the quietest account in the room to carry the same weight as the most senior one, when the most senior one usually speaks first and frames everything after it.

None of that is a failure of goodwill. It is what the working week is designed to produce.

Getting it to the people inside it

A pattern named at organisational level changes nothing on its own. It has to reach the people who are running it, in a form short enough to be read before a meeting rather than filed after one. That is what the Team Action Planning reports are for. Each team receives two pages: how the wider organisation experiences it, where the relationships work, the behavioural patterns its own data points to, and where to start.

The report is not the intervention. It is what makes the conversation possible, because it puts a shared account on the table that nobody has to be persuaded of first.

What the conversation is actually for

A shared account that arrives without a conversation becomes one more document. The obvious next move is to convene the team and agree the actions, which is the move that failed on the previous three attempts.

The value of the session is not the list of actions that comes out of it. It is the collective understanding that has to precede any action worth taking. A team that arrives at a task list early has converted a pattern back into a to-do, and a to-do cannot change the conditions that made the pattern reasonable.

Two disciplines make the difference. The first is collective responsibility, which in practice means each person describing their own part rather than diagnosing everybody else's. The second is holding both possibilities at once: when two teams score the same relationship differently, the instinct is to establish which of them is correct, but both are reporting their own experience accurately, and the distance between them is not measurement error. It is the finding.

Conclusion

Action still matters, and it belongs at the end rather than the beginning. One step, the one most within the team's own gift, changed deliberately; the next measurement then indicates whether it moved. Start with the actions and the pattern survives them. Start with the pattern and the actions have somewhere to land.

This is a slower path than data, insight, action. It is also the one that changes behaviour.