Trust is a system output, not a value you can put on a poster

Almost every organisation we work with has trust somewhere in its values. It is on the wall, in the induction pack, in the behaviours framework. In almost none of them has writing it down made any difference at all.

That isn't because the people involved are cynical, or because the value was badly worded. Trust is an output, the result of a system that either produces it or does not, mostly through mechanisms nobody has consciously designed.

You cannot declare an output. You can only change the thing producing it.

What is actually happening when trust is low

Here is a pattern we see repeatedly in high-pressure services: a team where trust is genuinely low, and where every individual in it is entirely trustworthy. Nobody is lying. Nobody is briefing against anyone. And yet nobody quite believes what they are told.

When you look at what is actually going on, it is almost always structural rather than personal. Decisions get made somewhere the team cannot see, and arrive as announcements. Someone raises a concern and it disappears: not rejected, just gone, with nothing coming back. A rota changes without explanation. A promise made in a meeting gets overtaken by events and is never mentioned again.

None of these are betrayals. Each one is a small prediction that failed to hold.

Trust breaks in small increments, and it is almost never about integrity.

This matters because of where teams look when they notice the problem. The research on organisational trust has distinguished its components for thirty years: broadly, whether someone is capable, whether they have your interests at heart, and whether they are honest.^[1] When a team says trust is low, they almost always reach for the third one. Someone must be being dishonest.

In our experience that is rarely where the damage is. The team has usually stopped being able to predict what will happen, and has reclassified an unpredictable system as an untrustworthy one. Those feel identical from the inside. But they need completely different responses.

Why the poster does nothing

Declaring trust as a value asks people to feel something. The system that produced the problem is still running, unchanged, and it will keep producing the same output. Worse, the distance between the stated value and the lived experience is itself corrosive: the team now has evidence of the gap between what it was told to expect and what it actually experiences. It also has evidence that leadership either cannot see the problem or has decided to name it rather than fix it.

We have watched organisations run trust workshops in teams where the actual issue was that nobody knew who made the decisions. The workshop is aimed at the wrong layer.

What changes when you treat it as an output

You stop asking people to trust each other more, which is not a thing anyone can do on request, and start looking at what the team is doing that makes trust hard to sustain. That is a considerably less comfortable conversation, because it usually points at how the team is led rather than at how its members feel.

It is also the only version of the conversation that changes anything.

About the author