Executive Coaching: What It Is, How It Works, and How to Evaluate One

Executive coaching is a contracted, one-on-one relationship between a coach and a leader, built around a defined set of goals and a fixed period of time, aimed at how that leader thinks, decides, and behaves in their role — not general life coaching, and not therapy for a mental-health condition. Most engagements exist inside a three-party arrangement: an employer or HR function is paying for it and has its own interest in the outcome, which means the leader in the room is not the only stakeholder. Who is paying, who receives what information, and who set the goals decide more about whether a given engagement actually works than any single coaching technique does — which is why this page starts there instead of with a model or a method.

Three parties, and only one of them is coached

The International Coaching Federation's own Code of Ethics, effective April 1, 2025, defines two separate roles inside a coaching engagement. The Client is the individual, team or group member being coached; the Sponsor is the entity or individual paying for, arranging or defining the coaching services. When someone hires and pays a coach for themselves, those two roles collapse into one person. In most executive coaching, they don't: the company, an HR function, or the leader's own manager is the sponsor, and the leader being coached is the client — two parties with two different sets of interests, having what looks like one conversation.

That split is why a serious executive-coaching contract has to say, in writing and before the first session, what the sponsor is entitled to know. The same code requires a coach to "maintain the strictest level of confidentiality with all parties involved," and separately requires a coach to have "a clear agreement about what information is exchanged and how it is exchanged among all parties involved during all coaching engagements" — which means the question of what the sponsor hears has to be answered explicitly, not assumed by either side. In practice, that's three concrete questions worth asking before signing anything: does the sponsor see attendance only, aggregated themes, a written progress report, or nothing beyond confirmation the engagement is running? Can the engagement be renewed, extended, or ended on the basis of something the leader never saw? And who decided the goals — the leader, the sponsor, or both together? A coach with specific, rehearsed answers to those three has actually thought through the arrangement they're in.

That three-party structure isn't universal, and it's worth naming the option that removes it entirely. IX Coach, an AI coaching system, is something a leader can use on their own, with no HR sponsor in the arrangement. That is a description of the arrangement, not a claim that removing the sponsor produces a better conversation: a sponsor in the room is exactly what makes some engagements work, and where that's true is covered later on this page.

What an engagement actually looks like

A competent engagement opens with a contracting conversation, ideally with the sponsor and the leader in the same conversation rather than two separate ones, and the point of that meeting is to turn something vague — "become a stronger leader" — into two or three goals specific enough that everyone in the room would recognize whether they'd been met six months later. A concrete goal names a behavior and who would notice it: delegating a defined category of decision rather than delegating more in general, changing what data reaches a particular stakeholder, or rehearsing a specific conversation the leader has been avoiding. Vague goals are usually a sign the contracting conversation didn't happen properly, not a sign the leader lacks direction.

Many engagements also draw on structured feedback from the people around the leader — peers, direct reports, a manager, sometimes the sponsor — gathered through an instrument or through interviews the coach conducts and synthesizes. This is genuinely useful raw material, and it's worth knowing what the research says about the feedback step in isolation, covered below: it is a data-gathering step, not the intervention itself. Collecting it and having the coach turn it into a working plan with the leader is where actual change gets built.

Sessions themselves are usually weekly or every other week across several months, structured as a conversation rather than an update — the coach asking rather than advising, often organized around a simple sequence like Goal, Reality, Options, Will (the GROW model), or built around Socratic-style questioning that leaves the leader owning the answer rather than the coach's advice. Between sessions there is normally something concrete to try and report back on, which is what keeps six months of conversation from becoming six months of describing the same problem in new language.

Explore: delegation levels · the grow coaching model · socratic method · effective one on ones

What the research actually supports

One academic answer to "does executive coaching work" isn't about executives specifically — it's a 2014 meta-analysis by Tim Theeboom, Bianca Beersma, and Annelies van Vianen in The Journal of Positive Psychology, pooling workplace coaching studies across performance and skills, well-being, coping, work attitudes, and goal-directed self-regulation. Their own conclusion: "coaching has significant positive effects on all outcomes with effect sizes ranging from g = 0.43 (coping) to g = 0.74 (goal-directed self-regulation)." Read plainly, those are moderate-to-large effects by the usual conventions for this kind of research — a real signal, not a rounding error.

It's also worth reading what the same paper says about the state of the field before its own results: "whereas coaching is very popular as a management tool, research on coaching effectiveness is lagging behind," and the studies available "generally lack a firm theoretical foundation." That's the researchers' own framing, not a hedge added by this page — coaching in general, and executive coaching specifically, is a field where the popularity of the practice has run ahead of the research documenting exactly why it works and for whom. A leader deciding whether to invest a leadership-development budget in coaching is on reasonably solid ground doing so; a leader expecting a number as clean as the effect sizes above to describe their own specific engagement is asking more of one meta-analysis than it can deliver.

Explore: self concept clarity

Where the line to a licensed clinician sits

Coaching, in the International Coaching Federation's own words, is "partnering with clients in a thought-provoking and creative process that inspires them to maximize their personal and professional potential." That definition is forward-looking and assumes a functioning person on the other side of the table — someone who wants clarity on a decision, a sharper way to lead, or momentum on a goal, not someone in crisis.

The practical line is the same one that applies to burnout, career stress, or any other workplace-adjacent coaching: if exhaustion or flatness follows the leader into the parts of life that have nothing to do with work, if sleep or appetite has changed, or if the conversation turns toward thoughts of self-harm, that belongs with a licensed clinician, not a coach — and in a crisis, an emergency service; in the U.S., calling or texting 988 reaches the Suicide & Crisis Lifeline at any hour. A coach, human or AI, who tries to coach through those signals instead of naming them and referring out is doing the wrong job with the right intentions.

Where a human executive coach is plainly the better choice

Some of what a good executive coach does in the room has no text-based substitute yet: reading the pause before a difficult answer, noticing what changed in a leader's posture when a particular name came up, sitting across the table during an actual conflict rather than hearing about it afterward. A board dynamic, a leadership transition, or a peer relationship that's quietly breaking down is read in real time by a perceptive person, and that is not a description of a technique — it's a description of a skill that currently only exists in a human being.

There's also a narrower, more procedural reason a company buying coaching often wants a human: accountability it can point to. "Coach" is not a licensed title in the United States, but the ICF's own credentials — ACC, PCC, MCC — require coach-specific education hours, logged coaching-experience hours, mentor coaching, and a performance evaluation or credentialing exam, and they can be checked through ICF's own coach-verification directory. A vendor contract can name a required credential level, and a procurement or legal function can verify it. An AI coaching product such as IX Coach carries no credential of that kind for a contract to name.

And the three-party arrangement described earlier on this page is, for some purposes, not overhead — it's the point. A coach who can sit in a mid-point review with the sponsor, translate what the organization actually needs into goals the leader owns, and be replaced if the fit is wrong is doing something a subscription a leader buys alone cannot do. If the engagement exists because of a succession plan, a return to a role after a serious failure, or a restructuring where multiple stakeholders need to see the same person credibly handling it, a credentialed human coach inside a real sponsor relationship is the honest recommendation, cost included.

How to evaluate an executive coach before you sign

Cost is a real signal, and the coaching industry's own numbers are a better starting point than either side's marketing. The ICF's 2025 Global Coaching Study — a survey with 10,035 valid responses from coach practitioners and managers using coaching skills, run by PwC Research — puts the global average fee for a one-hour coaching session at $234. Fees rise sharply with the generation of coach most likely to be senior and specialized: Baby Boomer coaches charge $270 an hour on average against $193 for Millennials, and coaches with ten or more years of experience average $69,721 a year in coaching revenue. That same experience bracket is also where executive specialization concentrates — 38% of coaches with a decade or more of experience primarily serve executives, against 18% of coaches with under a year in the field. A quoted rate well outside that range, in either direction, is worth asking about directly.

Beyond price, four questions do most of the work. What's the actual credential — ACC, PCC, MCC, or none — and can it be verified independently? What does the sponsor receive, in what form, and can you see a sample report before agreeing? What's the process for surfacing feedback from the people around you, and is it a real synthesis or a form nobody reads? And what does the coach do the moment the conversation turns clinical — do they have a named referral path, or do they keep going?

One caution on the feedback step specifically: a 2005 meta-analysis in Personnel Psychology by James Smither, Manuel London, and Richard Reilly, pooling 24 longitudinal studies of multisource ("360-degree") feedback, found that rating improvements over time were, in the authors' own account, generally small — collecting feedback is not, by itself, a reliable performance intervention. Ask what a coach does with the feedback once it's gathered, not only whether they gather it; the synthesis and the follow-through are where a 360 process either becomes useful or becomes paperwork.

Explore: the growth edge training

Where IX Coach fits

IX Coach sits outside the three-party structure described earlier on this page, because a leader can use it on their own. That is a real simplification for the everyday reps of leadership — thinking a decision through the night before it matters, rehearsing a hard conversation, coming back to a specific commitment between one week and the next — available at whatever hour that work actually shows up, drawing on relevant detail from earlier conversations rather than starting blank, and working from published concept pages such as the GROW model that a reader can open first. It is an AI coaching product, not a human coach, a clinician or a crisis service, and it does not replace what a sponsor relationship, a real 360 process, or a credentialed human coach's presence in the room provide.

Frequently asked questions

Research

• International Coaching Federation, (2025), ICF Code of Ethics, coachingfederation.org — Source for the Code's separate definitions of 'Client' and 'Sponsor,' the confidentiality standard (2.1) requiring the strictest level of confidentiality with all parties involved, and Standard 2.2 requiring a clear agreement about what information is exchanged and how among all parties involved. Standard 2.3, on disclosure to the appropriate authorities, is not used. Current version effective April 1, 2025. verified 2026-09-20
• International Coaching Federation, (2026), About ICF, coachingfederation.org — ICF's own statement of its definition of coaching, the wording quoted in the Where the line to a licensed clinician sits section. The page is not machine-fetchable by the quote gate; the wording was read directly from the page.
• PwC Research (for the International Coaching Federation), (2025), 2025 ICF Global Coaching Study: Executive Summary, International Coaching Federation — Sixth edition of ICF's own research into the coaching profession, 10,035 valid survey responses from coach practitioners and managers/leaders using coaching skills, collected Feb-Apr 2025. Source for the $234 global average hourly fee, the $270-vs-$193 hourly gap between Baby Boomer and Millennial coaches, the $69,721 average annual revenue for coaches with 10+ years' experience, the 38%-vs-18% gap in how often experienced vs. new coaches primarily serve executives, and the 'partnering with clients...' definition of coaching. Downloaded directly and text-extracted from the PDF rather than taken from a secondary summary. verified 2026-10-08
• International Coaching Federation, (2026), Credentials and Standards: ACC, PCC, MCC, coachingfederation.org — Source for the credential requirements named in the page (education hours, coaching-experience hours, mentor coaching, performance evaluation or credentialing exam) and for the Verify a Coach directory.
• Theeboom, T., Beersma, B., & van Vianen, A. E., (2014), Does coaching work? A meta-analysis on the effects of coaching on individual level outcomes in an organizational context, The Journal of Positive Psychology, 9(1), 1-18 — Source for the effect sizes (g = 0.43 to g = 0.74 across five outcome categories) and the authors' own framing that 'research on coaching effectiveness is lagging behind' how popular coaching has become. Verified against the published abstract reproduced on Vrije Universiteit Amsterdam's institutional repository page, since the publisher page is paywalled past the abstract. verified 2026-10-08
• Smither, J. W., London, M., & Reilly, R. R., (2005), Does performance improve following multisource feedback? A theoretical model, meta-analysis, and review of empirical findings, Personnel Psychology, 58(1), 33-66 — Source for the caution that rating improvement across 24 longitudinal studies of 360-degree feedback was generally small. Verified against a summary that reproduces the study's own abstract and citation. Used only for the qualitative finding ('generally small'), not for a specific effect-size number, since that number could not be independently confirmed against a source this page could read directly.