The 2% Problem: Why Coaching Works but Almost Nobody Gets It.
When I took the uniform off for the last time, after thirty-one years in the Italian Navy, I found myself doing a strange inventory. What exactly had the Navy taught me?
Quite a lot, as it turned out. I had learned how to bring several thousand tons of steel alongside a pier without doing expensive damage. I knew how to navigate, plan operations, work with intelligence, understand military law, manage communications, and follow the procedures that keep people alive when the sea becomes unpleasant.
For almost every skill expected of a naval officer, there had been a course, an instructor, an exam, and eventually somebody signing a piece of paper saying I knew what I was doing.
Leadership was different.
Over thirty-one years, including years commanding people in difficult environments, very little of my formal training focused on what would eventually matter most: what happens when other human beings depend on you.
Since then, I've spent much of my professional life working as a coach, trying in various ways to fill that gap for other people. That eventually led me to an uncomfortable question about coaching itself.
My new book, The Last Human Coach, is built around that question.
Before I go any further, let me disclose that I co-founded a company, Besage.ai, that builds iLeader, an AI coaching system. I'm interested in this subject, so please make whatever adjustment you think is needed. I would rather put that on the table now than have you discover it later.
Coaching works, for real.
If you work around coaching long enough, you eventually encounter the claim that coaching returns five, six, sometimes seven times what it costs.
The figure most often quoted is 5.7 to one. It comes from a 2001 study involving 100 executives. Only 43 provided a monetary estimate of the return they believed coaching had generated, and those estimates were retrospective and self-reported.
The study isn't useless, but I would be reluctant to walk into a CFO's office waving “5.7x ROI” around as though it were a law of physics.
There is better evidence. A 2023 meta-analysis looked only at randomized controlled trials. The researchers analyzed 39 coaching samples involving 2,528 participants and found a statistically significant positive effect on leadership and personal outcomes. The estimated effect was moderate. They also found publication bias, and the estimate became smaller after adjusting for it.
I find that reassuring, because it still proves that coaching works and has a meaningful impact.
But... who gets it?
There aren't enough coaches
Let's do some math. The 2025 ICF study estimates roughly 123,000 professional coaches worldwide, about 110,000 of whom have active clients. On the other side of the fence, the U.S. Bureau of Labor Statistics counts more than 11.1 million people in management occupations in the United States alone.
Imagine every active coach in the world stopped working with everyone else and served only managers in the United States. Each coach would still have roughly one hundred managers to look after. And nobody anywhere else would have received a single session yet.
Calculating exactly what percentage of managers currently receive professional coaching is messy, but we can make a rough estimate. Using ICF's 2023 North American numbers, you get roughly 429,000 active coaching relationships. About 56% of practitioners said their clientele consisted mostly of managers or executives.
Use that as a rough proxy, and you arrive at around 240,000 manager-or-executive coaching relationships across North America. When you compare that with 11.1 million U.S. managers, the rough calculation comes out to about 2%.
The numbers are not perfectly aligned. The numerator covers North America while the denominator is U.S. managers, and “mostly managers” isn't the same as counting every individual client. Even if the exact percentage is somewhat higher or lower, the underlying problem doesn't disappear.
Professional coaching reaches only a small proportion of the people responsible for other people's working lives.
I call this the 2% problem.
The rest are the neglected 98%.
Human attention is expensive
The obvious explanation for the existence of the Neglected 98% is money. According to ICF, the average coaching fee is about $234 for a one-hour session, so a twelve-session engagement costs $2,808.
If you gave those twelve sessions to every one of the 11.1 million U.S. managers, your bill would rise to roughly $31.2 billion.
For context, the entire global coaching profession generated an estimated $5.34 billion in annual revenue.
Scarcity is the reason coaching is expensive. A coach may work with twelve or thirteen active clients at a time. Their attention is limited, and every hour spent with one person is an hour that cannot be spent with another.
So organizations make choices, and guess who is more likely to get a coach.
Senior executives, and maybe high-potential employees.
The rest, the neglected 98%, when they are lucky, get the consolation prize: a workshop.
We are investing at the wrong end of the organization
This becomes particularly strange when you look at what we know about managers. Gallup has reported that managers account for at least 70% of the variation in employee engagement across business units. For most employees, the daily experience of the organization is shaped by the person they report to, who decides how meetings feel, how mistakes are handled, whether effort gets noticed, whether a disagreement becomes a conversation or a small war.
Yet individualized leadership development tends to become scarcer the closer you get to the front line. The people with the most direct contact with employees often get the least support. The problem gets bigger if you consider that many of them became managers because they were excellent engineers, salespeople, analysts, technicians, or project managers, not leaders. Promotion arrived as recognition for being good at one job and quietly changed the job into something else.
Gallup has estimated that only about one in ten people naturally possesses a high level of managerial talent. That leaves a very large number of people learning leadership while already responsible for a team.
Workshops are being asked to do too much.
I run workshops, and I know how good workshops can be, but they happen at a particular moment, while the difficult conversation they address may happen six weeks later.
A few years ago, I had dinner in Rome with an old high-school friend. His company had sent him to a huge leadership event the year before, one of those with a campus speaker and plenty of positive energy. So, I asked what he remembered from it and, after a moment of silence, he said: “I walked on hot coals.”
That was it! A year later, the thing that had survived was the firewalking. I am not mocking him or the event. Peak experiences are memorable and energizing. But I am also sure his employees did not need him to remember walking barefoot over hot coals.
They wanted him to become slightly better at managing them, repeatedly.
That is a different kind of development.
I know what one bad manager can cost
I once served under an officer who was professionally competent in almost every conventional sense. He was also the worst leader I ever worked for. He shouted constantly, and nothing was quite right unless he had done it himself. People learned to spend much of their energy anticipating what would make him angry.
Now, consider that I had wanted to join the Navy since I was a child. During those two years, for the first time in my life, I wanted to leave. One badly equipped human being came closer to pushing me out of a profession I loved than any enemy I ever encountered (and, BTW, I wasn't the only one in that miserable situation).
That experience changed the way I think about leadership development.
A manager's lack of skill travels and changes how people sleep. It changes what they say in meetings. It changes whether they volunteer an idea or keep quiet. Eventually it changes whether they stay.
Gallup has found that 42% of employees who voluntarily left an organization believed their manager or organization could have done something to prevent their departure.
Turnover is expensive. Replacing a frontline employee may cost around 40% of salary. Replacing a manager can cost roughly twice their annual salary.
The quieter cost shows up in the people who remain but slowly decide to care less.
So what reaches the other 98%?
If coaching works, and if the traditional model is inherently limited by the number of human coaches available, then training more coaches only takes us so far, because human attention and availability are finite.
So what happens to the manager sitting at home at eleven o'clock at night, knowing that tomorrow morning they have to tell a good employee that their performance isn't good enough?
What happens to the person who has just become a manager and realizes that one member of the team openly resents their promotion?
What happens five minutes after a meeting goes badly, when the manager knows they mishandled it but doesn't yet understand how?
Those are coaching moments that most probably will never happen in front of a coach.
So, I spent a book trying to answer this question: how can we bring some of what coaching does well much closer to the moment people actually need it?
The Last Human Coach: How to Bring Coaching to Every Manager will be out on September 28, 2026, only on Amazon!
Comments