Quick answer
The most quoted figure, a sevenfold return, comes from a single 2009 survey. The International Coaching Federation commissioned it and PricewaterhouseCoopers carried it out with the Association Resource Centre. Companies self-reported the number, and most could not produce the figures behind their answer.
Stronger evidence exists. A 2023 meta-analysis of 37 randomised controlled trials found a moderate effect on performance and behaviour. It offers no dollar multiple. It does survive scrutiny.
What produces return is the design of the engagement. A baseline taken at the start. Agreement between coach, leader and sponsor. Commitments tracked between sessions. A review against that first baseline. This article sets out which claims hold up, and how to build an engagement that produces evidence a board can read.
You already know something needs to shift.
Maybe you are a senior leader making decisions nobody else in the building fully understands. Maybe you are the HR or L&D head trying to justify a six-figure coaching investment to a CFO who speaks only in spreadsheets. Either way you have landed on the same question. Is executive coaching worth the money?
If you are building that case, you have already met the seven-times figure. It sits in almost every coaching brochure.
The number is real in one narrow sense. Someone reported it. Nobody measured it.
That gap is the subject of this article.
What does “worth it” actually mean?
Before the numbers, the question needs defining. “Worth it” means different things to different people.
For the executive, it means: will this make me sharper in my decisions and more effective with the people I lead? Will I stop repeating the patterns holding me back?
For the HR director or sponsor, it means: can I show a return? Will this move retention, engagement, and the strength of the pipeline in ways I can report upward?
Both questions have answers. The harder question has never been whether coaching works. It is whether your specific engagement is built to produce anything you can point at.
If you would rather talk through your own situation than read the studies, book a conversation.
Where does the 7x ROI figure come from?
One study, published in 2009.
The International Coaching Federation commissioned it. PricewaterhouseCoopers and the Association Resource Centre carried it out. Companies reported a median return of seven times their investment, and 86% said they had at least made their money back. Individual clients reported a median of 3.44 times.
You will often see those two figures quoted as though they corroborate each other. They do not. They come from the same survey.
Two things are worth holding onto. Respondents were largely the trade body’s own members, so people who already believed in coaching. And most could not produce the figures behind the number they gave.
That does not make it fabricated. It makes it a marketing statistic rather than a research finding, and a seventeen-year-old one at that.
Which ROI claims survive scrutiny?
Put the industry number and the academic evidence side by side and the difference is not subtle.
| A buyer would ask | The “7x ROI” survey | The controlled research |
|---|---|---|
| Who ran it? | The coaching industry’s trade body, with PwC | Independent university researchers, peer reviewed |
| Was there a control group? | No comparison with people who had no coaching | Yes, with random assignment |
| Who answered? | Largely the trade body’s own members | Participants across independent organisations |
| Could respondents back up the number? | Most could not produce the underlying figures | Full results published, 2,528 people, 39 samples |
| What is the actual finding? | A self-reported financial multiple | A measured, significant gain in performance and behaviour |
| Would it hold up if a sceptical board member checked? | No | Yes |
If a CFO pulls one thread, the survey figure comes apart. The second column does not.
What does the controlled evidence show?
Strip out every methodological weakness critics raise about coaching research. Self-selection. No control group. Self-report bias. Look only at randomised controlled trials, the design used to test new medicines.
The effect holds.
A 2023 meta-analysis in Academy of Management Learning & Education pooled 37 randomised controlled trials, 39 independent samples, 2,528 participants. It found a moderate effect on performance and behaviour, statistically significant at p < .001.
Translated for a general reader, the average coached leader ended up performing better than roughly seven in ten comparable leaders who had no coaching. That translation is the standard way of expressing an effect size of g = 0.59. The paper reports the effect size. It does not state the seven-in-ten figure itself.
The authors also flag some indication of publication bias. Worth knowing if you plan to use this as proof that coaching always works. It supports a narrower claim. Coaching holds up under scrutiny.
An earlier meta-analysis, Theeboom and colleagues in 2014, found coaching’s largest measured effect was not on how people feel. It was on their ability to set a goal and follow through (g = 0.74). Larger than its effect on wellbeing, work attitudes, or coping.
Neither study hands you a dollar figure. Both would survive a board member checking the source.
What does the payback actually look like?
Abstract ratios are useful for approvals. Concrete arithmetic is better.
Take a single retention save. Gallup puts the cost of replacing a leader at around twice their annual salary. That counts recruitment, lost output during the vacancy, and ramp time. If a coached leader keeps one senior person who would otherwise have gone, that avoided cost covers the coaching several times over.
The larger returns are harder to put on a spreadsheet. They show up in decisions made under pressure. In conflicts resolved instead of escalated. In a team that works better because the person running it does.
Often the pattern looks like this. A leader six months into a bigger role, attrition climbing, two direct reports raising the same concern about how they communicate. Nothing has broken yet. The numbers move before anyone calls it a problem.
The framing that matters is not “what did I spend”. It is “what is different now, in this leader and in their team, because this happened”.
Does coaching improve performance and productivity?
There is one study people reach for here, and it is almost always misquoted.
Olivero, Bane and Kopelman studied 31 managers in a US public agency in 1997. The same managers went through conventional management training, then eight weeks of one-to-one coaching, with productivity measured after each stage. It rose 22.4% after training and 88.0% after coaching.
There was no control group and no random assignment. The 88% is therefore a cumulative figure for a single cohort, not the amount coaching added on top of training. You will usually see it reported as the gap between two groups. There were no two groups. One agency, 31 managers, nearly thirty years ago. Treat it as suggestive, not settled.
The 2009 ICF client study found 80% of coaching clients reported improved self-esteem or self-confidence, and more than 70% reported better work performance, relationships, and communication. Self-reported again, and about clients generally rather than executives specifically.
How does coaching affect engagement and retention?
Here is a number that should concern any leadership team.
Global employee engagement fell from 23% to 21% between 2023 and 2024. Manager engagement fell further, from 30% to 27%. Gallup put the productivity cost at US$438 billion and identified managers as the driver of the decline.
Managers are also where coaching does its most direct work.
The 2023 ICF and Human Capital Institute study Defining New Coaching Cultures found 72% of respondents saw a strong link between coaching and employee engagement. That is a perception measure from an industry study, so hold it the way you would hold the 7x figure.
Why coaching rather than training, mentoring, or a 360?
Coaching is not the only leadership development tool, and it is not always the right one.
| Instead of coaching | What it does well | Where it stops |
|---|---|---|
| Mentoring | Transfers hard-won experience from someone who has done the job | Tends to transfer the mentor’s approach rather than build the leader’s own |
| Training programmes | Delivers knowledge and shared language at scale | Knowing what to do and doing it under pressure are different problems |
| Peer advisory groups | Perspective and accountability from people at the same altitude | Breadth over depth; rarely gets at the specific pattern in the way |
| 360-degree feedback | Shows you where the gaps are, with evidence | Tells you the gap exists; does not close it |
The Olivero finding above is the clearest argument for combining them. Training alone moved productivity. Training plus coaching moved it further.
The best engagements use 360 data as a baseline and build the work around what it shows.
Coaching outperforms the alternatives when it is the right intervention, for the right person, at the right time. That is what a first conversation establishes. No pitch, no obligation.
Where does coaching fail?
Not all coaching delivers any of this. The evidence about where it goes wrong is clear enough.
The coach is underqualified
The industry has no universal certification requirement. Harvard Business Review flagged that over two decades ago and it is still true. A coach working with senior leaders needs deep professional training and real business judgement. Then the psychological literacy to know when they are out of their depth. An ICF credential is a baseline signal. It says nothing about whether someone has led anything.
There is no outcome framework
Coaching without defined objectives is an expensive conversation. What does success look like at three, six and twelve months, and how will you know? Without that scaffolding the return disappears, because nobody set up anything to measure.
The executive is not ready
Coaching works when someone will examine their own patterns and assumptions. If they arrive wanting validation rather than challenge, it stalls. I have said no to executives who only wanted to tick the attendance box.
The organisation does not support it
Coaching in a vacuum rarely sticks. When performance systems, team dynamics and leadership expectations all pull the same way, the return compounds. Without that, even good coaching hits a ceiling.
Who gets the most from executive coaching?
Four situations, consistently.
Leaders in transition. New C-suite appointments, first-time people leaders stepping up from technical roles, executives taking regional or global scope for the first time. This is where the work shows fastest, because the behaviour has not set yet.
High-potential leaders being built for bigger roles. The earlier someone develops self-awareness and stakeholder judgement, the longer that compounds across a career.
Leaders under pressure. Burnout, decision fatigue, stakeholder conflict, the isolation of senior roles. These are performance risks, not welfare issues. Deloitte’s 2025 Global Human Capital Trends work argues that as AI reshapes the leadership role, coaching and development capability matters more rather than less.
Organisations going through change. Mergers, restructures, market pivots, rapid growth.
A note on that last one. You will see it claimed that 60 to 70% of change initiatives fail without coaching. Do not use that figure. Mark Hughes examined the five most-cited sources for it in the Journal of Change Management in 2011 and found none of them carried empirical evidence. Each either asserted the number or cited someone else who had. If a vendor quotes it at you, that tells you something about the vendor.
How do you evaluate whether coaching is right for you?
Ask a narrower question
Not “is coaching worth it” in the abstract. Is coaching the right intervention for this challenge, this person, this moment?
Vet the coach
Look for accreditation, then look past it. Ask about their experience with leaders at your level facing your kind of problem. Ask how they measure outcomes. Ask what they do when the coaching is not working. The answer to that last one tells you the most.
Define success first
Agree it between coach, coachee and sponsor before you start. What outcomes matter, how will progress be tracked, what does good look like. This is not bureaucracy. It is the only thing that makes a return measurable later.
Commit to the process
Behaviour change appears with consistent sessions and clear goals, and how quickly depends on the situation. The compound effect builds across the engagement. Coaching is not a quick fix.
The bottom line
The honest position is narrower than the brochures and stronger than the sceptics expect.
There is no reliable dollar multiple. The figures that offer one come from surveys of people with an interest in the answer, most of whom could not show their working.
Coaching produces a measurable improvement in how leaders perform and behave, demonstrated in randomised controlled trials and published in peer-reviewed journals. That claim will hold when a board member checks the source.
The return then depends on things you control. A baseline. A qualified coach. An agreed definition of success. An organisation that lets new behaviour survive contact with the calendar.
The real risk is not the money. It is leaving good leaders without the thing that would make them, and everyone reporting to them, measurably better.
If you want to work out whether that applies to your situation, thirty minutes is enough to find out.
Frequently asked questions
Does executive coaching actually work?
Yes, by a tougher standard than the ROI surveys use. A 2023 meta-analysis in Academy of Management Learning & Education pooled 37 randomised controlled trials covering 2,528 participants and found a moderate effect on performance and behaviour. Randomised trials are the design used to test new medicines, though the authors report signs of publication bias and note that self-reported outcomes show larger effects than observed ones. It gives no dollar figure.
Is the 7x ROI figure for executive coaching real?
It is a real survey result, not a research finding. It comes from one 2009 study commissioned by the International Coaching Federation and carried out by PricewaterhouseCoopers and the Association Resource Centre. Respondents were largely the trade body’s own members, there was no control group, and most could not produce the figures behind the number they reported. Treat it as a marketing statistic.
How do you measure the ROI of executive coaching?
Take a baseline before the engagement starts, then measure against it afterwards. We use the Leadership Circle Profile 360 for that baseline, because it measures behaviour against a normed reference group rather than asking people how they felt about the coaching. Other useful markers include engagement survey results, retention data, and whatever performance indicators the role already reports. Pair those with qualitative feedback from the coachee, their stakeholders, and the sponsor. The engagements that produce a defensible number are the ones that defined it at the outset.
How much does executive coaching cost?
Fees vary widely with the coach’s experience, the structure of the engagement, and the seniority of the leader. The Clarity Practice publishes its own fees on its service pages rather than quoting an industry range that would tell you little. The more useful question is what the engagement is designed to change, and how you would know whether it had.
How long does an executive coaching engagement last?
Most engagements run from three months to a year, with sessions typically every two to four weeks. How quickly behaviour change shows depends on the situation and on the work done between sessions. Some organisations keep ongoing coaching relationships for their most senior leaders.
What is the difference between executive coaching and mentoring?
A mentor shares advice and experience from having done the job. A coach works to build the leader’s own capacity to think, decide and act differently. Mentoring tends to transfer the mentor’s approach. Coaching develops the leader’s own. Both have value, and they answer different problems.
Is executive coaching only for struggling leaders?
No. Coaching high performers who want to operate at the next level is a long-established use, and it is where much of the corporate spend goes. Remedial coaching exists, but treating coaching as a remedial signal misreads what it is for.
Corrections
9 September 2026: a further correction to the Olivero, Bane and Kopelman finding. The study had no control group and no random assignment: the same 31 managers were measured after training and again after coaching. Earlier versions of this article, including the 26 August correction below, described a coached group as distinct from a trained-only group. No such comparison exists in the study. The effect size from the 2023 randomised-trial meta-analysis is also restated as moderate rather than moderate-to-large, and the authors’ reported publication bias is now noted.
26 August 2026: an earlier version of this article reported the Olivero, Bane and Kopelman productivity finding as a difference between coached and trained-only managers. It is the gain within the coached group. That figure and the comparison now match the study.
