Ask a capable leader, in a calm moment, how to handle a struggling team member, and they will tell you fluently: listen first, separate the behaviour from the person, be specific, agree clear next steps. They know it cold. Then sit in on the actual conversation, on a Tuesday, when they are behind and stretched, and watch them do almost none of it — faster, more controlling, reaching for the familiar move instead of the right one. This is the knowing–doing gap, and it is the most reliable pattern in leadership development. It is also the most misunderstood, because almost everyone treats it as a gap in knowledge, and it is not.
The instinctive response to a leader who isn't leading well is to send them on something — a course, a programme, a new framework. It rarely moves the needle, and there is a structural reason. A leader with a richer vocabulary doesn't get more action; they get a more sophisticated way to describe the action they didn't take. The new knowledge converts into a better post-hoc explanation of the gap rather than into closing it. You have almost certainly seen this directly: the manager who has read everything about feedback and gives none of it, now able to explain, in well-sourced detail, why this particular situation wasn't right for it.
This should be fatal to the assumption underneath most development, and it usually isn't. The assumption is that knowing leads to doing — that the journey from understanding to action is short and automatic. In reality, knowledge was never what produced action in the first place. Watch anyone perform well under pressure in any field and the knowledge has disappeared into the performance. The surgeon isn't reciting anatomy; the experienced negotiator isn't running through a checklist. What produces action in the live moment is not what a person knows. It is who they are being.
Under pressure, a leader doesn't reach for what they know. They express who they are being — and if the person they collapse into under stress is a controlling, threat-managing one, then all the good knowledge is sealed off behind a state that cannot reach it. Take a worked example. A director needs to give a capable but defensive manager some hard feedback. She knows exactly how to do it. But on a bad day she walks into the room being someone for whom this conversation is a threat to be contained — and from inside that, what shows up to her is not a colleague to be levelled with but a problem to be managed. The feedback she knows how to give is simply not available to her; what is available is a softened, hedged version that protects everyone from discomfort and changes nothing. Same knowledge, different person in the chair, completely different conversation.
Now suppose she has done some work on who she is being, and walks in as someone for whom the conversation is an act of respect she owes a colleague. The same knowledge is suddenly reachable, because the situation she now inhabits is one in which candour is the natural move rather than a risk to be managed. Nothing changed in what she knew. Everything changed in what was available to her to do.
The leader who can describe great listening at 9am interrupts everyone by 3pm. The knowledge didn't leave. The conditions changed, and it became unreachable.
This also explains why the gap is invisible in training and obvious in the work. A classroom gives a leader time to consult what they know deliberately, under no real pressure. The Tuesday afternoon removes the time and applies the pressure, and tests who they actually are when there's no space to think. Development that only adds knowledge is, almost by definition, rehearsing for conditions that never occur. It is why so many programmes earn glowing feedback forms and produce no change a month later: they worked perfectly on the layer that wasn't the problem.
If the gap isn't made of missing knowledge, more knowledge won't cross it. Two things do. The first is working on who a leader is being — not adding to what they know, but shifting the observer they operate as, so that effective action becomes available under load instead of sealed off behind a defensive state. This is harder than content and cannot be delivered as content, which is precisely why it's under-done. The second is making the concrete moves trainable: how to make a clear request, secure a real commitment, and close the loop. Most of what gets labelled an execution problem is a broken move at exactly this level, and these moves can be practised.
For anyone responsible for developing leaders, the practical takeaway is a better question to ask of any programme. Not “what will our leaders learn?” — that is satisfied by a good set of slides — but “what will our leaders be able to do, under pressure, that they can't now, and how will we see it in how they actually behave?” A programme that can only answer the first question will reliably produce more articulate, no more effective leaders. The second question is the only one that closes the gap.
We go deeper on the mechanism, with a three-layer model of where the gap opens and what closes it, in the research paper The Knowing–Doing Gap in Leadership.