Why Most Executive Coaching Fails to Change Behaviour

An executive walks out of a coaching engagement able to name exactly what has been holding them back. They can describe the pattern, the trigger, the cost to their team. Six months later, a colleague who worked with them before the coaching would struggle to tell the difference. The insight was real. The behaviour never moved.

I have sat on both sides of this problem for more than 20 years, first as a senior leader leading organisational transformations across the UAE and globally, and then as an ICF Professional Certified Coach (PCC) working with C-suite leaders across the UAE, the wider GCC and internationally. The pattern above is not the exception. It is closer to the norm, and it explains why so many capable executives quietly conclude that coaching did not work for them.

Why Executive Coaching Fails to Change Behaviour

Executive coaching fails to change behaviour most often because the engagement ends the moment insight is reached, before the new behaviour has been tested, reinforced and observed by the people around the leader. Clarity and change are two different events. An executive can understand precisely what they need to do differently and still walk back into the same meetings, under the same pressure, and default to the same reflex, because insight alone does not compete with twenty years of habit.

Coaching that stops at awareness is doing half a job and calling it complete.

Insight Is Not the Same as Change

Most coaching conversations are genuinely good at producing insight. A skilled coach asks the right questions, holds up a mirror the executive rarely sees in daily life, and helps them connect a leadership pattern to its actual cost. That moment of recognition feels significant, and it is. It is also frequently mistaken for the finish line.

I have worked with senior leaders who could describe their own patterns with striking precision. One executive I coached could explain exactly how his need to control every decision was slowing his team down and quietly training them to stop bringing him anything that required real judgement. He understood it intellectually within the first two sessions. Delegating differently took him closer to a year, and only started moving once we built specific, observable behaviours into how he ran his weekly leadership meetings, not once we simply discussed the pattern further.

Understanding a pattern and interrupting it under pressure are two different skills, and executive coaching that only builds the first one will keep producing executives who can describe their blind spots beautifully and still act on them the same way next quarter.

The Coaching Stops the Moment the Hard Part Starts

This is where most executive coaching engagements quietly fail. The behavioural change research on this point is some of the clearest in the entire leadership development field. Marshall Goldsmith and Howard Morgan's widely cited study, drawn from over 86,000 survey responses across more than 11,000 managers in eight companies, found that the single strongest predictor of whether a leader's effectiveness actually improved was not the quality of the coaching or training itself. It was whether the leader kept following up with colleagues afterward, asking for input and showing visible, ongoing effort to change, as Goldsmith and Morgan's research on the follow-up factor demonstrated.

That single finding reframes what executive coaching is actually for. The sessions themselves are the setup. The follow-up, the accountability, the repeated small corrections over months, is where behaviour actually changes. A coaching engagement built around sessions alone, with no structured mechanism for what happens between them, is optimised for insight and largely blind to the part of the process that determines whether anything changes.

Nobody Around the Leader Is Told What Is Changing

A related failure shows up almost as often. An executive works hard in coaching, genuinely shifts a behaviour, and returns to a leadership team that has no idea anything is different. Peers still brace for the old reaction. Direct reports still filter what they say based on how the executive used to respond. The executive tries the new behaviour once or twice, gets no signal that it landed, and quietly reverts, not because the change failed, but because nobody around them was told to expect it or reinforce it.

Coaching that stays entirely confidential between coach and executive, with no visibility for the manager or the wider team, makes this outcome more likely, not less. Confidentiality about the content of sessions is essential to any coaching relationship. With the executive’s agreement, relevant stakeholders can be made aware of the specific behaviours being developed without compromising the confidentiality of the coaching conversations.

The Wrong Problem Gets Coached

Executive coaching also fails when it treats a systemic problem as a personal one. A leader gets coaching for poor delegation, when the real issue is that the organisation has never clarified decision rights, so nobody below them is actually authorised to decide anything meaningful without escalating it. A leader gets coaching for communication style, when the real issue is an organisational structure that forces five approval layers between an idea and a decision, and the leader's abruptness is a symptom of operating inside constant time pressure, not a personality flaw.

Coaching the individual in these situations produces a more self-aware executive still trapped inside an unchanged system. The behaviour that looked like the problem was never fully theirs to fix alone.

What the Research Actually Says About Coaching That Works

It is worth being direct about the numbers here rather than either dismissing coaching or overselling it. The International Coaching Federation's Global Coaching Client Study found that a large majority of client organisations, 86 percent, reported at least a full return on their coaching investment, with individual clients reporting a median return several times the cost of the engagement. That data point matters, but it also masks a wide range of outcomes underneath it. Coaching that includes structured follow-up, clear success measures agreed before the engagement starts, and some level of manager or stakeholder involvement consistently outperforms coaching run as a purely private, session-only relationship.

The gap between those two versions of coaching is not really about coach skill. It is about whether the engagement was designed, from the outset, around the part of the process where behaviour actually changes.

What Makes Coaching Actually Change Behaviour

A few design choices consistently separate coaching that produces lasting behaviour change from coaching that produces a more articulate version of the same executive.

Behaviour gets named specifically, not generally. "Communicate better" cannot be practised. "Ask two questions before giving my opinion in the Monday leadership meeting" can be, and can be observed by the people in that room.

Someone besides the coach is watching for the change. Whether that is a manager, a peer, or a structured 360 process built into the engagement, behaviour that only the coach hears about rarely gets reinforced anywhere else.

The engagement runs long enough to survive pressure. Genuine behaviour change tends to hold under calm conditions early and then gets tested hard the first time real pressure returns. Coaching engagements that end right before that test happens miss the moment that would have proven whether anything actually changed.

Follow-up is built in, not left to willpower. Bi-weekly sessions across six to twelve months, with clear checkpoints for what has actually shifted in observable behaviour, give change a far better chance of surviving than a short, front-loaded burst of sessions.

How to Know If a Coaching Engagement Is Built to Fail

A few warning signs tend to show up early. The engagement has no defined success measure beyond the executive reporting they found it valuable. Nobody besides the coach and the executive knows what specific behaviour the coaching is targeting. The engagement is scheduled to end at a fixed date regardless of whether the new behaviour has actually been tested under pressure yet. And the organisation treats the coaching itself as the intervention, rather than one part of a wider system that also needs to notice and reinforce the change.

Any one of these on its own is worth addressing before the engagement starts. Several of them together usually predict the outcome I described at the start of this piece: real insight, and very little that a colleague would actually notice six months later.

Coaching Does Not Fail. Coaching Designed Around Insight Alone Does

Executive coaching is not the weak link most people assume it is when a previous engagement did not stick. In my experience, a well-designed coaching engagement has a far greater chance of shifting behaviour when the leader is genuinely willing to do the work. What I have seen, repeatedly, is coaching designed around a single hour every two weeks, with no follow-up structure, no visibility for the people around the leader, and no plan for the moment pressure returns and old habits reassert themselves.

Insight is the easy part of this work. It is also the part most coaching engagements are accidentally optimised for. The harder, less visible work, the follow-up, the reinforcement, the repeated small corrections over months, is where behaviour actually changes, and it is the part that determines whether an executive walks away from coaching with a better vocabulary for their patterns or an organisation that genuinely experiences them differently.