Why Does Leadership Behaviour Revert After Executive Coaching Ends?

Sep 09, 2026

Quick Answer

Behaviour reverts because most coaching engagements work at the level of knowledge and capability, while the behaviour itself is generated somewhere beneath it. Under pressure the leader does not lose the skill. They lose access to it. Until the levels that produce the behaviour are reached, insight holds for as long as conditions stay calm, and stops the week they do not.

You watched the leadership behaviour change hold for about eleven weeks.

The coaching was good. You chose the coach carefully, the 360 feedback was accurate, and the leader engaged with it properly rather than performing engagement for your benefit. Something moved. You saw it in the first month. So did their team.

Then the quarter turned. The board asked a hard question, the leader was in front of it without preparation time, and the person you had sat across from in that July session was back in the room by the second week of October, behaving exactly as they had before anyone spent a dollar on the engagement.

You are not looking at a failure of diligence. You commissioned the assessment, you selected the provider, you had the conversation that most people in your position postpone for another cycle. Nothing in that sequence was wrong. The change never reached the level where the behaviour is produced.

Why does the change hold for a while and then stop?

Because behaviour is state-dependent. Most development work is not.

A leader who has understood something about themselves in a calm room, with a skilled practitioner, no audience and no consequence attached to the next sentence, has acquired something real. What they acquired is available in conditions resembling the ones they acquired it in. Change the conditions and the availability changes too. The same executive can describe their pattern with real precision on Tuesday and enact it without noticing on Thursday.

That is not hypocrisy. It is not weak commitment either. It is access.

You don't lose the skill under pressure. You lose access to it. The knowing-doing gap that every L&D function has watched open between a well-received programme and the following quarter is not a motivation problem, and treating it as one produces a second intervention that fails for the same reason as the first.

What level was the coaching actually working at?

Almost certainly the top one. The surface.

Most executive coaching, most 360 debriefs and most leadership programmes operate at the level of knowledge and capability. That level is necessary. A leader who cannot name what they do under pressure cannot change it, and insight of that kind is genuinely valuable. It is not where durability comes from.

Beneath capability sit four further levels, and behaviour is generated by all of them working together:

Beliefs are the rules the leader is operating from, most of which were formed long before they had a title and none of which are examined during a competency conversation. A belief that visible uncertainty is dangerous will quietly overrule a communication framework every time.
Values determine what the leader will protect when two good things are in conflict, which is the situation that produces almost every behaviour a 360 flags.
Identity is who the leader believes they are, and it is the level at which most change quietly fails, because a person will consistently act in ways that keep their sense of themselves intact even when it costs them.
The nervous system decides, faster than deliberate thought, whether the current moment is safe enough to do the new thing. A workshop does not persuade it.

An intervention that works at the first level and stops there is not a bad intervention. It is an incomplete one. The reversion you saw at week eleven was predictable rather than surprising.

Is this a discipline problem or a systems problem?

It is a systems problem. That distinction is worth defending when the next leadership budget request comes back to you.

The most common organisational response to reversion is to conclude that the leader lacked follow-through, and the second most common is to conclude that the coach was not strong enough. Both conclusions are available. Both feel reasonable. Both send you back to the market for a slightly better version of the intervention that already did not hold.

The more useful reading is narrower. Regulation manages the pattern; resolution changes why the pattern exists. A leader who has learned to steady themselves before a difficult meeting has acquired something worth having, and they will still be steadying themselves before that same meeting in three years, because the reason the meeting activates them has not been touched.

What should you check before commissioning the next intervention?

Five questions. Ask any provider before a contract is signed.

What level does this work at, and how would we know? Ask the provider to name it. A provider working at capability will say so plainly if asked directly, and that answer is useful rather than disqualifying.
What happened to the last three engagements at six months? Not at completion. At six months, under the pressure that generates the behaviour in the first place.
Has this specific behaviour resisted a previous intervention? If it has, the question is no longer which provider is better. It is which level the previous attempt did not reach.
What does the provider do when the insight lands and the behaviour does not follow? Most methods have no answer to this, and the answer is where the real difference sits.
Would the leader's own team be able to tell the difference in ninety days? If nobody outside the coaching room could detect the change, the organisation has bought insight rather than behaviour.

That's the work I do: I develop leaders, and certify the coaches who develop them, at the four levels where change actually holds — nervous system, identity, beliefs, and values. That's the Meta-B Leadership System.

What to take from this

Reversion after a well-run engagement is a signal about level, not about the leader's character or the provider's competence.
The knowing-doing gap is a question of access under pressure rather than a question of motivation.
Capability work is necessary and insufficient; the durability sits beneath it.
Before commissioning the next intervention, establish what happened to the provider's previous engagements at six months rather than at completion.
If a behaviour has already resisted one serious attempt, the second attempt should be aimed at a different level rather than at a better version of the same one.

On Thursday 17 September at 11:00 AST I'm running a seventy-minute live demonstration called Below the Surface, locating where a real leadership pattern is generated across five levels. It's a demonstration of diagnosis, not of intervention. A second sitting runs on Thursday 24 September.

Bring one leader in mind.