What is executive coaching? In my practice it is a structured, confidential working relationship between a senior leader and a coach, built around real decisions the leader has to make, with a commitment to act before the next session. It is not therapy, mentoring or training, and the research says it works, with a moderate effect, though not for every leader or every engagement.
Ask ten people what executive coaching is and you will get ten versions of the same vague sentence: a professional development process where a coach helps a leader grow. That sentence is true and it is useless. I run these engagements for a living, and the version I give a board is more specific: a recurring conversation that pressure-tests a live decision, ends with something the leader will do, and starts the next session by checking whether they did it.
I built the Capability Forum, and the 5D Leadership Model, because I got tired of watching leaders pay for coaching that was really just expensive company. If your coach never disagrees with you, never assigns you homework and never asks what you did between sessions, you do not have a coach. You have a listener with a day rate.
What is executive coaching, structurally?
Structurally, executive coaching is three things on a repeat cycle: a real decision on the table, a structured conversation that pressure-tests the leader's thinking on it, and a commitment the leader is held to before the next session. Take one away and you have something else wearing coaching's clothes.
- A live decision, not a theme: Every session starts with something concrete the leader is actually deciding: who to promote, whether to restructure a team, how to handle a peer who's undermining them. Not 'let's talk about delegation this month.' Themes are for workshops. Coaching runs on decisions.
- Structured pressure, not sympathy: The coach's job in the room is to ask the question the leader has been avoiding. Not to validate the plan they walked in with. If every session ends with the leader feeling reassured rather than slightly uncomfortable, the coach isn't doing the job.
- A commitment with a deadline: Every session ends with something specific the leader will do before the next one, and the very first thing the next session covers is whether they did it. No commitment, no accountability, no coaching: just conversation with an invoice attached.
What executive coaching is not
The confusion comes mostly from what coaching gets mistaken for. It sits next to three other things people conflate it with, and the differences are not subtle once you look at direction and intent rather than format.
- Not therapy: Therapy looks backward: why do you feel this way, where did this pattern start. Coaching looks forward: given where you are, what's the next right move. A good coach will refer a leader to therapy when the conversation needs to go backward. That's a boundary, not a failure.
- Not mentoring: A mentor draws on their own path and tells you what they'd do in your seat. A coach doesn't need your job history to be useful: the value is in the rigour applied to your specific decision, not in shared war stories.
- Not training: Training delivers the same content to everyone in the room on a schedule set by the curriculum. Coaching has no curriculum: the agenda is whatever decision is live that week, which is precisely why it doesn't scale the way training does, and precisely why it works when training alone doesn't.
The same distinction applies when people ask me to compare executive coaching with leadership consulting: consulting hands you an answer built from outside your context, while coaching builds the answer with you, inside it, and then makes you own the decision. If the label is what confuses you, executive coaching versus leadership coaching covers that boundary.
How an executive coaching engagement works, step by step
In my practice an engagement runs as a set cycle, not an open-ended series of chats. Strip away the marketing language the industry hides behind and the shape looks like this.
| Stage | What happens | Typical timing in my practice | What you should have afterwards |
|---|---|---|---|
| Contracting | Agree the focus, how progress will be judged, who sees what, and what confidentiality means if your organisation is paying | Before the first session | A written agreement, including what a sponsor is told and what stays private |
| Diagnostic | One session, sometimes supported by short conversations with the people who work with you, to find the two or three decisions or behaviours that matter most now | First session | A short list of priorities, not a personality profile |
| Working sessions | Every two to three weeks. Each opens with what you did with what we agreed last time | Months one to three of a cycle | Decisions made differently and written down |
| Midpoint recalibration | Stop and ask whether we are still working on the right thing, because the second problem is usually more honest than the first | Around the midpoint | A reset agenda and an honest view of progress |
| Close and handover | Name what changed and what you now do differently without me in the room | End of the cycle, usually three to six months; CEO engagements often run six to twelve | A record of the changes and a plan for sustaining them |
Weekly sessions sound more supportive. In my experience they mostly produce conversation without enough happening in the real world between sessions to make the next conversation useful. The gap between sessions, where a hire actually gets interviewed and a difficult conversation actually gets had, is where coaching either worked or did not.
How strong is the evidence that executive coaching works?
I would rather give you the evidence and its limits than a brochure. The strongest recent evidence is encouraging and modest, and it comes with caveats a buyer should know.
- A 2023 meta-analysis by de Haan and Nilsson covered 37 randomised controlled trials of workplace and executive coaching, with 39 samples and 2,528 participants. It found a statistically significant effect, with a best estimate of g = 0.59 that the authors call moderate, and it warns of significant publication bias. Effects were larger when outcomes were self-reported than when they were observed, and they did not depend much on the length of the assignment.
- A 2023 meta-analysis by Cannon-Bowers and colleagues, covering 11 studies after removing an outlier, found an average effect of g = 0.44 and concluded that coaching is an effective workplace intervention. It found no significant difference between face-to-face and virtual coaching, and neither the number of sessions nor the total hours of coaching significantly predicted outcomes.
- A 2009 review by De Meuse and colleagues found a moderate-to-large gain in skill or performance, but also that people who receive coaching report stronger effects than others do, and that the return from coaching was inconsistent and in some circumstances could be negative.
Taken together I read the evidence as: coaching works on average, the effect is moderate, and the average hides a lot of variation. Two practical points follow. First, because self-reported outcomes run higher than observed ones, ask what the people around you would notice, not only how you feel. Second, because hours and session counts did not predict results, do not buy on volume. Buy on whether the engagement is tied to decisions and judged against objectives agreed at the start.
The market is also large and still growing, which is part of why quality varies. The International Coaching Federation's 2025 Global Coaching Study, conducted by PwC, reports 122,974 coach practitioners worldwide and an estimated $5.34 billion in revenue over the past year, and 73 percent of coaches agree that clients and organisations expect a coaching credential. Credentials are a reasonable filter, but the Cannon-Bowers authors note that very few studies report the coach's certification, so the research cannot yet show that credentials improve outcomes. For a practical filter, see how to choose an executive coach.
My own lens: Capability, not comfort
Everything above is fairly standard practice among coaches who take the discipline seriously. Where I differ is what I am building toward. Most coaching frameworks aim for the leader to feel more confident. I aim for the organisation around them to be more capable, which is the premise of Leadership Capability Architecture: the leader is not the finished product, the system they build and run is.
That is also where reciprocal leadership comes in. A leader who can only make good decisions when I am in the room has not been coached. They have been rented a crutch. The test I use throughout an engagement is whether the people around the leader are getting better at their jobs because of how the leader now operates. If the leader improves but the team does not, the coaching has stayed personal instead of becoming structural, which is the gap the 5D Leadership Architecture exists to close.
That is why I do not run coaching as a fixed weekly ritual disconnected from the business. The cadence bends to the decision, not the calendar. A leader mid-restructure needs a different rhythm than a leader who has just closed their toughest quarter. A tidy recurring slot suits the coach's diary, not the leader's outcome.
Who executive coaching is actually for
The assumption I hear most is that coaching is for leaders who are struggling. The pattern that matters more is the reverse: leaders who are winning can carry the best-hidden blind spots, because results cover the gaps. A failing leader gets the truth handed over by circumstances. A succeeding leader has to go looking for it, because nothing in their environment volunteers it.
The leaders I see getting the most from it tend to fall into one of four groups.
- Leaders stepping into a materially bigger remit than the one that got them promoted
- Founders making the specific transition from doing the work to running the people who do the work
- Leaders whose team, on a blunt read, would describe them as harder to work for than the results justify
- Leaders who are succeeding by the numbers but privately unsure why, or unsure it will hold
If your real question is whether it is worth the money, I have set out the numbers and their limits in the ROI of executive coaching. For how long results take, see how long executive coaching takes to show results. And if the situation is not a coaching problem at all, when executive coaching is the wrong intervention explains how to tell.
What your organisation should agree before the first session
If a company is paying, three conversations should happen before the coach and the leader meet. The first is with the sponsor, usually the CEO, chair or CHRO, about what would count as a good outcome in twelve months. The second is about confidentiality: what the sponsor will hear, in what form and with the leader's knowledge. The third is about the exit. Every engagement should say how either party can end it, and what happens to the fee if the fit is wrong.
I put these in writing because most coaching disappointments I see trace back to a gap in one of them, not to the sessions. A sponsor who expected a behaviour change and a leader who expected a private sounding board will both leave unhappy, and neither is wrong about what they wanted. Writing it down turns a vague hope into something you can review at the midpoint. It also protects the leader. Coaching only works if the leader can say what is actually going on, which requires knowing in advance where the edges of the conversation are.
The honest test of whether coaching is working
I give clients one test to run on their own engagement, with me or with anyone else. Can you name, specifically, what you did differently in your last three real decisions because of it? Not what you learned. Not how you feel. What you did. If the honest answer is that you are not sure, the engagement has drifted into company rather than capability. That can still be pleasant, even valuable as a sounding board, but it is not what it was sold as.
One more thing to ask any prospective coach before you sign: what will you hold me accountable for, and how will you know if I did not do it? If the answer is vague, the engagement will be vague. If it is specific, you are talking to someone who understands what executive coaching is, not someone selling the comfortable version of it.
Coaching is also no substitute for the leader doing the work between sessions. Any coach who implies a leader can fix a problem without changing their own behaviour is selling something other than coaching. For chief executives specifically, see what makes executive coaching for CEOs effective.
If you want to look at your own situation, the work with me page sets out where I start, and CapabilityAI, my AI leadership companion for the weeks between conversations, is described on its own page.