Executive coaching is a confidential, time-limited working relationship in which a senior leader examines how they think, decide, and lead, with a trained practitioner whose only agenda is the leader's effectiveness. It is not advice, not training, and not treatment, though it borrows from each. A typical engagement lasts three to twelve months, meets roughly twice a month, begins with a precise question, and ends with a change that other people can see.
I am asked the question in the title more often than any other, usually by someone who has been told to find a coach and is not entirely sure what they are shopping for. The confusion is reasonable. The phrase covers an enormous range of practice, from a weekly call with a retired chief executive who tells war stories to a structured engagement run by a clinician with assessment instruments and a written plan. Both call themselves executive coaching. They have almost nothing in common.
This essay is an attempt at precision. I want to say what the work is, what it is not, who it is for, and what a well-run engagement actually involves, so that a leader considering it knows what they are buying.
A Working Definition
Strip away the marketing and executive coaching has four defining features.
First, it is a relationship between two people, one of whom holds significant responsibility for others. The subject of the work is that person: their judgment, their habits of attention, their way of handling pressure, their effect on the people around them. The company is context, not client, even when the company pays.
Second, it is confidential. The leader must be able to say what they actually think, including things they cannot say to the board, the team, or their spouse, or the work does not happen. I have written separately about what actually stays private, because the details matter more than the promise.
Third, it is bounded in time and organized around a question. Open-ended coaching with no end point and no stated purpose is companionship, which has value but is not this. A real engagement starts by naming what should be different at the end of it.
Fourth, the practitioner's only interest is the leader's effectiveness. A coach is not selling a software implementation, not angling for a board seat, not reporting to HR about fitness for promotion. That independence is the whole source of the work's value, and the first thing to verify when choosing a coach.
What It Is Not
The boundaries are as important as the definition, because most disappointments in coaching come from engaging the wrong instrument.
It is not mentoring. A mentor has done your job and tells you how they did it. That is valuable and it is also limited, because their path is not yours and their advice carries their blind spots. A coach does not need to have run a company to help you run yours, any more than a physician needs to have had every disease.
It is not consulting. A consultant studies the business and recommends what to do. A coach studies the leader and helps them see what they are doing. If the problem is the strategy, hire a consultant. If the problem is that a sound strategy keeps failing in execution because of something about how you lead, that is coaching.
It is not training. Training transfers a skill to a group. Coaching works on one person's particular pattern, which is usually not a missing skill but a habit that was once useful and now is not.
It is not therapy. This is the boundary I am most often asked about, and the one that matters most. Coaching assumes a well person pursuing better performance. Therapy treats distress. The difficulty is that burnout, anxiety, depression, and the long tail of sleep deprivation and alcohol very often present at work as leadership problems: indecision, irritability, withdrawal, a loss of the judgment that got someone the job. A coach without clinical training will work on the leadership problem for months while the actual condition entrenches. I have set out where the line falls and how to tell which side you are on.
Who It Is For
The honest answer is: people whose decisions carry consequences for many others, and who have few places to think without a political cost. Chief executives, founders, the C-suite, senior partners, the general counsel, the leaders being readied for those roles. The common thread is not the title. It is exposure.
Within that group, the people who get the most from coaching share a few traits. They came voluntarily, or at least arrived willing. They can tolerate hearing something unwelcome about themselves. And they have a specific question, even if it is only half-formed: why do I keep hiring the same kind of person, why does the board relationship feel managed, why am I so much worse at this than I was two years ago.
The people who get the least are those who were sent as a formality, those whose real problem is structural rather than personal, and those who want a witness rather than a change. Coaching can still be useful in those cases, but the expectations should be set accordingly.
What an Engagement Actually Involves
Practice varies, so I will describe mine, which is representative of the more structured end of the field.
It begins with a conversation about whether coaching is the right instrument. Roughly a third of the time I conclude that it is not, or not yet, and say so. Of those who proceed, the first two or three sessions are assessment: how this person thinks, what their history is, what the people around them experience, and whether anything clinical is sitting underneath the presenting problem. Sometimes that includes structured feedback from colleagues; I have written about what 360 feedback shows and what it misses.
From there, the work is a sequence of conversations, usually every two weeks, each organized around what happened since the last one and what the leader intends to do differently before the next. In between, the leader runs experiments: a different way of opening a board meeting, a delegation they would normally not make, a conversation they have been avoiding. The sessions examine the results. Over months, the pattern that brought them in becomes visible, then optional, then gone. I describe the shape of this in more detail in what happens in a coaching session.
Engagements run three to twelve months. Shorter ones address a single defined problem, a transition or a specific relationship. Longer ones take on something more structural about how a person leads. The right length is a question in its own right.
What Makes It Work
The research on coaching outcomes, which I have reviewed elsewhere, is clear on one point: the method matters much less than the conditions. Voluntary participation, a specific goal, a strong working relationship, and a correct diagnosis of the problem predict success. The brand name of the framework does not.
That last condition, correct diagnosis, is where I would put the emphasis for anyone deciding whether to engage. The most expensive failure in coaching is not a bad coach. It is a competent coach working diligently on the wrong problem because they lacked the training to recognize what was actually in front of them. The capacity to tell a workable pattern from a treatable condition is not a nice-to-have in this work. It is the thing that determines whether the months are well spent. I have written about what clinical training changes in practice, and about how it shapes this practice specifically.
If you are considering coaching and want to test whether it is the right instrument for what you are facing, the first conversation is free and confidential, and I will tell you honestly if it is not.