Most coaching conversations end with something that sounds like a plan and behaves like a wish. I’m going to be more open with Daniel. Two weeks later, nothing has happened, and neither of you can say exactly what should have.
How do you work on trust between coaching sessions?
You don’t work on trust. You practise one behaviour, with one person, in situations you can name in advance.
Trust is not a thing you can act on directly. It is someone’s confidence in you with something they care about — their conclusion, drawn from what they have experienced of you. You cannot reach in and adjust it. What you can change is the evidence they are drawing from, and that changes one observable action at a time.
So the between-sessions question is never “how do I improve trust with Daniel”. It is: which single thing will I do differently, with Daniel, and when will I know the moment has arrived?
Nothing changes in the session. The session is where you decide what to change; the weeks in between are where the other person finds out.
Why do the weeks between matter more than the hour in the room?
Because the other person is not in the room.
Whatever insight you reach in a session is invisible to them. They were not there, they did not hear it, and their view of you will not move by a millimetre until something they can observe is different. The session produces a decision. Only the following fortnight produces evidence.
That is uncomfortable for a reason worth naming: it means the value of coaching is realised somewhere the coach cannot see, on days nobody is watching, in moments that last thirty seconds.
What exactly should you practise?
One behaviour, chosen because it answers the question that is actually unresolved for that person.
If you have made a Trust Map, you already have this: the largest gap on something that carries weight for them points at a focus area, and the focus area offers a small number of behaviours that would strengthen it. Pick one. Not three.
One is not a limitation, it is the design. Three behaviours practised inconsistently produce a person who seems slightly different in ways nobody can name. One behaviour practised consistently produces something the other person could describe out loud — and being describable is what makes it evidence.
Why does “I’ll be more open with Daniel” fail?
Because it names an aspiration and not a moment, so there is never a point at which you are visibly not doing it.
The most useful finding here is about the shape of the plan rather than its content. Peter Gollwitzer and Paschal Sheeran pooled 94 independent tests of what happens when people convert a goal into an if-then plan — naming the when, the where and the how in advance. The effect on whether the goal was actually reached was medium-to-large. Holding the intention on its own did considerably less.
The practical translation is to write the trigger, not the value:
- Not “be more open with Daniel” → “When I spot a slip on Thursday’s build, I’ll message Daniel that day rather than waiting for Monday’s stand-up.”
- Not “listen better” → “In our Tuesday one-to-one, before I respond to a concern, I’ll say back what I think they meant.”
Those studies are mostly health and achievement behaviours, in laboratories and field trials — not trust between colleagues. The mechanism is what travels: a plan attached to a situation gets acted on more reliably than a plan attached to a virtue. Do not read the effect size across to your Thursday.
How often should you practise it?
Little and often, rather than a push in the week before the next session.
Two separate bodies of evidence point the same way. On learning, Nicholas Cepeda and colleagues synthesised 839 assessments across 317 experiments: spacing repetitions out beat cramming them together, and the ideal gap grew the longer the material had to last. On trust specifically, a 2026 meta-analysis of repeated trust games found that how consistently a partner reciprocated was by far the strongest influence on trust — stronger than the size of any individual act.
Neither of those gives you a schedule, and anyone who tells you it does is overreaching. The spacing work is word lists and paired associates; the trust work is laboratory games with money. What both support is a direction — frequency and consistency do more than intensity — and that direction argues against the thing most people actually do, which is to remember the commitment the night before the session.
What do you bring back to the session?
Evidence, not a report card. Three things are worth writing down as they happen, because none of them survives two weeks of memory:
- What you did, and when. The specific instance, dated. “Flagged the schedule risk on the 6th, before the review.”
- What they did. Their response, as neutrally as you can put it. Not what you think it meant.
- What you avoided. The moment you saw, recognised, and did not take. This is usually the most useful line on the page, and it is the one people leave out.
What if nothing has changed?
Often nothing has, and that is information rather than failure.
There are three honest explanations and they lead to different places. The evidence hasn’t accumulated yet — a fortnight of one behaviour is a small amount of new information against years of the old, and consistency is exactly the thing that takes time to read. You practised it and they didn’t see it — the behaviour was real but landed somewhere they weren’t looking, which is worth checking directly. Or you worked on the wrong question, which happens more than anyone expects, because the question you find easiest to work on is rarely the one they are unsure about.
What none of the three supports is doing more, harder. Escalating effort on a behaviour that is not the one in doubt produces a person who is trying visibly and still not trusted, which is worse than where you started.
What between-sessions work is not
It is not homework, and framing it that way makes it worse. Homework is completed for the person who set it. This is practised for the person it is about, who never sees the coaching and is the only one whose view counts.
It is not a substitute for the session either. The decision about which question is unresolved, and which behaviour would answer it, is genuinely hard to make alone — the whole difficulty is that people are poor judges of which of the five they are weakest on. The session is for choosing. The fortnight is for evidence. Neither does the other’s job.