When leaders ask me how to choose an executive coach, I tell them to check three things before the first call: whether the coach has led at your level, what they are credentialed in, and whether their process is written down. Then run the test that matters more than any of those. Will this person challenge your thinking, or just validate it? Most coaches, even good ones on paper, default to validation. It is comfortable, it renews the contract, and it feels like support. It isn't coaching. It's expensive company.
I've sat across the table from a lot of executive coaches, both as a client early in my own career and now as the person leaders call when a previous coaching relationship went nowhere. The pattern is depressingly consistent. The coach who gets rehired every quarter is rarely the one who moved the leader furthest; they're the one who made the leader feel good in the room. Those are different skills, and only one of them is worth paying for.
How to choose an executive coach: The filter that actually matters
Every generic guide on this topic gives you the same three filters: check their credentials, check the chemistry, check the price. None of those predict whether the engagement will change anything. Two filters provide a better test, and any leader interviewing a coach should run both before signing anything.
The first filter: does this person have a repeatable framework, or are they improvising a conversation? A coach without a framework is reacting to whatever you bring into the room that week, which feels responsive, but means session six looks nothing like session one, and there's no way to tell whether you're actually progressing or just talking. A coach with a real framework (mine is built on the Leadership Capability Architecture model) can tell you, before you've paid a penny, what the arc of the engagement looks like and how they'll know it worked.
The second filter, and the one people skip because it's uncomfortable to test for: will they tell you when you're wrong? Ask any prospective coach to describe the last time they pushed back hard on a senior client and the client didn't like it. If they can't produce a specific example, they haven't done it; a coach who has never disagreed with a paying client is not disagreeing with you either, no matter how the sales call goes.
Seven checks before you sign
Filters are useful only if you can run them in a conversation. The table below turns the two filters above, plus the practical points most buyers forget, into seven checks you can put to any coach on a shortlist. Take it into the call and fill it in as they talk. A coach who answers six of the seven with specifics is worth a second meeting. A coach who answers three of them with general statements about trust and potential has told you how the engagement will feel, and not what it will change.
| Check | Where to look | A strong answer | A red flag |
|---|---|---|---|
| Level fit | Ask for their last three engagements, described without names | Roles, company size and decision stakes close to yours | They have coached leaders in general and cannot say at what level |
| Challenge record | Ask for the last time a senior client resisted their view | A specific, recent, slightly uncomfortable example | Talk of holding space, with no example |
| Credential | Ask for the credential level and check it with the issuing body | A named level and the hours behind it, offered without prompting | A long list of badges and no level you can verify |
| Written method | The proposal or engagement letter | Stages, outputs and review points on paper | 'It depends where you want to take it' |
| Measurement | How progress is recorded and who sees it | Indicators agreed in week one and reviewed at set points | Satisfaction scores and nothing else |
| Sponsor and confidentiality | The engagement letter | What the sponsor sees and what stays private, in writing | An understanding that is never written down |
| Exit | The contract | An end date or a review point, with a clean way to stop | An open-ended retainer with no review |
None of these checks needs inside knowledge of coaching. They need only that you ask and that you wait for the answer. If you are comparing named options in a particular city, my comparisons of executive coaching programmes in London, Sydney and New York describe each provider from its own published material, so you can run these checks against something written down.
What credentials tell you and what they do not
The International Coaching Federation sets three core credential levels. According to its credentials page, the ACC credential requires 60 or more hours of education and 100 or more hours of coaching experience. The PCC requires 125 or more hours of education and 500 or more hours of experience. The MCC requires a PCC credential, 200 or more hours of education and 2,500 or more hours of experience. Read as a ladder, that gives you a quick way to ask what a coach's letters mean in hours.
Credentials also matter to the market. The 2025 ICF Global Coaching Study, which received 10,035 responses from 127 countries, reports that 73% of coaches agree that clients and organisations expect them to hold a coaching certification or credential. That is coaches describing what clients ask for, so I treat it as a signal about expectations and not as proof of quality.
My own view is that a credential is a floor. It tells you a coach has logged supervised hours and met a published standard. It does not tell you whether those hours were spent with people carrying your level of responsibility, and it says nothing about whether the coach will disagree with you in week two. So I ask for the level, I check it with the issuing body, and then I move on to the questions that separate coaches who hold the same letters. If you are sitting on a board and weighing London candidates, how a board can run that selection is the next read.
Does an executive coach need to have been a CEO?
Not necessarily, and I would be wary of anyone who says yes without qualification. What matters is that the coach has carried accountability for decisions with real consequences and can recognise the pattern when you describe one. A former CEO brings pattern recognition and credibility in the first meeting. The risk is that they start advising, and advising is consulting, which is a different service with a different failure mode: you end up borrowing their judgement when the point was to improve your own.
For a CEO I would rather hire a coach who has held a P&L, a board relationship or a turnaround, and who will question my reasoning, than a former chief executive who tells stories from their own tenure. The practical test is to ask for the last engagement where the stakes were comparable to yours and to listen for what they did, not what they achieved. I go further into this question in what makes executive coaching for CEOs effective.
Validation versus challenge: The split that decides everything
Validation-mode coaching sounds like this: reflecting your own words back at you, asking open questions that lead nowhere in particular, and ending every session with some version of "it sounds like you already know the answer." Sometimes you do. Often you don't; you know the answer that's comfortable, and the coach just helped you feel confident about it instead of testing it.
Challenge-mode coaching sounds different. It names the pattern you can't see because you're inside it. It asks the question that makes you defensive, and then sits in the silence instead of rescuing you from it. In my own practice, the moment that tells me a coaching relationship is working comes when the client pushes back, hard, because I've said something that lands somewhere real. Their agreement tells me far less. Agreement is easy to manufacture. Friction that produces a better decision is not.
Why a repeatable framework beats improvised conversation
I built the 5D Transcending Leadership Model™ because I got tired of watching capable executives cycle through coaches who were, in effect, running an unstructured therapy session with a business vocabulary bolted on. A framework earns its place by letting you tell, in month three, whether you're actually further along than you were in month one, on a dimension you agreed mattered at the start. Making coaching feel more corporate is no part of its job.
Without a framework, every session resets. The coach asks what's on your mind this week, you talk about whatever fire is burning, and you leave with some perspective but no compounding progress. With a framework, each session builds on the last one against a fixed set of capabilities: judgement, influence, the ability to develop other leaders rather than just direct them, and so on. If a coach can't describe their framework in two sentences, they don't have one, no matter how many years they've been practising.
This is how I apply the Leadership Capability Architecture lens to the buying decision itself. The model rests on five structural pillars, and each one translates into a question you can put to a coach before you pay anyone a fee.
The five pillars as a buying checklist
- Alignment System: What has to change in the business, and who must agree it has changed? A coach should be able to restate your objective as a business outcome, not as a feeling you want to have.
- Leadership Identity System: Who has this coach actually led or advised? Look for evidence at your level and in your kind of company, not a general claim to have coached leaders.
- Execution Cadence System: What is the session rhythm, and what happens between sessions? Coaching that only exists in the room has nothing to compound.
- Capability Pathways: What is left behind when the engagement ends: documented decisions, a stronger bench, a way of working that the team can repeat?
- Cultural Reinforcement System: What will your board, sponsor and team see change, and how will they hear about it? Behaviour that nobody around you can see has not yet changed the organisation.
Reciprocal coaching, not top-down instruction
There's a version of executive coaching that's really just consulting with better bedside manner: the coach tells you what to do, you nod, you leave. That model fails for the same reason reciprocal leadership beats command-and-control inside organisations: the person closest to the problem (you) usually knows more about your context than the coach does. The coach's job isn't to hand down answers. It's to apply enough structured pressure that your own judgement, tested properly, produces a better answer than it would have alone.
That's a different working relationship than most people expect when they hire a coach. It means the coach will ask you to defend a decision rather than simply approve it. It means you'll leave some sessions more uncertain than when you arrived, because uncertainty that's honestly earned is more useful than confidence that was never tested.
Questions to ask before hiring a coach
- Describe the last time you told a senior client something they didn't want to hear.: If they can't give you a specific, recent, uncomfortable example, they either haven't done it or won't do it with you. Vague answers about "holding space for difficult conversations" are a tell, not a qualification.
- What's your framework, and what does progress look like on it by session six?: You want a concrete answer about what capability is being built and how it's measured, not "it depends on where you want to take it," which really means there's no structure at all.
- What happens when I disagree with your read of a situation?: Listen for whether they describe a genuine back-and-forth or immediately concede. A coach who folds the moment a senior client pushes back has already told you how the whole engagement will go.
- Who have you worked with that this didn't work for, and why?: Every coach with real experience has had engagements that failed. If they claim a perfect record, they're either not being honest with you or they've only ever taken clients who were never going to challenge them.
- How do you handle it when I'm the problem, not my team or my market?: Most executive problems presented as external turn out to be a leadership behaviour underneath. A coach who consistently helps you externalise blame is coaching your ego, not your capability.
- What does the engagement look like if it's working, versus if it's just pleasant?: Ask them to distinguish the two out loud. If they can't articulate the difference between a coaching relationship that feels good and one that's actually changing outcomes, they haven't thought about it, which means they can't tell which one you're in either.
- Who sees what I say, and where is that written?: You want a plain answer about what the sponsor receives and what stays between the two of you. If the answer is an assurance and not a document, ask for the document.
- What credential do you hold, at what level, and what hours sit behind it?: A good coach answers in one sentence and tells you where to check. Hesitation here is information, even if the coach is excellent.
Written terms, sponsors and confidentiality
Most executive coaching goes wrong in the contract, not in the room. Two people agree on a rhythm and a fee, nobody writes down what the work is for, and six months later each remembers a different promise. When you choose an executive coach, the engagement letter is where you find out whether they run a process or a series of conversations.
I would want five things on paper before the first session. The objective, stated as a change in the business or in how you decide, not as a general wish to improve. The cadence, including what happens between sessions. The duration, with a review point at which either side can stop without argument. The measures, meaning what you and the sponsor will look at to decide whether the work is moving. And the confidentiality terms: what the sponsor receives, in what form and how often.
Sponsors deserve a separate word. If your chair, your CEO or your CHRO is paying, they have a legitimate interest in knowing the work is real. They do not have a legitimate interest in the content of your sessions. The cleanest arrangement I know is to agree objectives and review points with the sponsor at the start and the end, share a short written summary of progress against the agreed measures in between, and keep the content of sessions between you and the coach. If you are self-funding, you can skip the sponsor, but write the same terms down anyway, because the discipline is for you.
Watch also for who owns the data. If the engagement includes a 360 assessment or psychometric profile, ask who sees the report, how long it is kept and whether you can see everything that is written about you. A coach who has thought about this will answer quickly. One who has not will say it has never come up, and that is also an answer.
What the research says about whether coaching works
The best recent summary I know is a 2023 meta-analysis by Cannon-Bowers and colleagues in Frontiers in Psychology. Across 11 studies it found an overall effect of coaching that was positive and moderate, with a Hedges' g of about 0.43 to 0.44 in its summary tables. The authors also say there is a lack of controlled studies and that several moderators rested on a single study, which limits how much weight any one finding can carry.
I read that honestly. Coaching in general has a reasonable evidence base for improving workplace outcomes. That tells you nothing about whether a particular coach will help you in particular, which is why the checks above matter more than any statistic about the profession. If you want the harder question of what return to expect, I have set out what research can and cannot show in the ROI of executive coaching.
What to budget
Fees vary by city and by seniority, and the figures below are published by providers, so they are asking prices and not independent market data. In London, one coaching firm's published guide puts most executive coaches at roughly £200 to £300 an hour, with a top tier at £350 to £500 an hour (Mary Taylor and Associates, 2026). In Australia, an Australian consultancy publishes a range of $300 to $900 an hour for an established coach and $900 to $1,500 or more for a C-suite specialist, excluding GST (M Kable Consulting, August 2026). My city comparisons above carry the detail. Use the numbers to spot an outlier in either direction, then judge the engagement on scope and not on rate.
What good executive coaching actually costs you
Not money. Attention. The real cost of good coaching is the discomfort of being told, by someone you're paying, that the story you've been telling yourself about a decision doesn't hold up. Leaders who choose a coach based on how comfortable the sessions feel are chasing the wrong measure. Comfort is what you get from friends. What you're paying a coach for is the willingness to be the person in the room who doesn't need you to like them in order to keep doing their job properly.
I've watched this play out the same way often enough to trust the pattern. A leader hires a coach who's warm, well-credentialed, easy to talk to. The sessions feel good. Six months in, the leader is more articulate about their problems but hasn't actually changed a single decision because of the coaching. That's not a coaching failure in the dramatic sense (nobody did anything wrong, exactly); it's a slow, expensive way of confirming what the leader already believed. The money bought reassurance, not judgement.
Contrast that with the engagements that actually move something. They tend to be less comfortable in the room and more useful outside it. The leader leaves some sessions irritated, occasionally defensive, and comes back the following month having actually changed how they ran a difficult conversation, or reversed a call they'd been too attached to admit was wrong. That's the trade you're making when you choose a coach who challenges rather than validates: less comfort now, more capability later. Most people, if they're honest, would rather have the comfort. That's exactly why so much executive coaching underperforms its price tag.
If you want a fuller picture of what separates capability-based coaching from the generic version, I've written about it directly in why your C-suite needs capability-based coaching: the short version is that most coaching treats symptoms in the room, while capability-based coaching builds the underlying judgement that prevents the next ten problems from needing a coach at all.
The one-sentence test
If you remember nothing else from this: hire the coach who makes you defend your thinking, not the one who makes you feel understood. Understanding is free: you can get it from your team, your partner, your friends. What you're paying an executive coach for is the discipline to keep pushing after everyone else in your life has learned to stop.
I put the distinction this way. A good coach changes how you decide. A great engagement changes how the company decides without you. The first is a conversation you pay for. The second is something you can still see in the way your team runs a meeting a year after the last session. If you are weighing a coach against a wider programme of leadership work, my ways of working with me page sets out where I start, and CapabilityAI, my AI leadership coach, is grounded in the same methodology. If you want to see where your team stands before you hire anyone, a leadership assessment gives you something concrete to bring to the coach.
For the CEO case in particular, see what makes executive coaching for CEOs effective.