The honest answer is: it depends on conditions most people don't think to ask about before they sign the engagement letter.
The research on executive coaching ROI is consistent and genuinely impressive. The International Coaching Federation reports that 86% of organizations that tracked their coaching investment at least broke even, with a median return of five to seven times what they spent. A Manchester study of 100 executives found an average return of 5.7 times the initial investment. MetrixGlobal reported 529% ROI, climbing to 788% when employee retention benefits were included. Those numbers come from real organizations measuring real outcomes, and they've been replicated enough times across enough contexts to take seriously.
But here's what those numbers don't tell you: they're measuring the average across a wide range of coaching engagements, good and mediocre, well-matched and poorly matched, with leaders who were ready to do the work and leaders who weren't. The average looks good because the high end pulls it up. The question isn't whether executive coaching works. It does. The question is whether it will work for you, in your specific situation, right now.
What the research says coaching actually changes
Seventy percent of people who go through coaching report improved work performance. Sixty-one percent report improved business management. Fifty-seven percent report better time management. Those outcomes show up consistently across studies, which suggests that good coaching reliably produces a certain set of results: better self-awareness, sharper decision-making, improved relationships with direct reports, clearer communication, stronger ability to delegate without abdicating.
What the research also shows, and what most coaching conversations skip, is that the quality of the coaching matters enormously. Tasha Eurich's research found that while 95% of people believe they are self-aware, only about 10% actually are. Coaching accelerates self-awareness faster than almost any other leadership development intervention. But only if the coach has the skill to surface what the leader can't see, and only if the leader has the honesty to look at what's surfaced.
The ROI surveys ask leaders who chose coaching whether it worked. That's a self-selected sample, people who were already motivated to engage. The real question isn't whether coaching works for people who are committed to it. It's whether the conditions for commitment are present in your situation.
When it's worth it
Coaching has the highest return when a few specific conditions are true. The leader is operating at a level where their decisions have meaningful leverage. A CEO whose judgment affects a hundred people or a hundred million in revenue gets more from coaching than someone whose decisions affect a smaller system, simply because the same improvement in decision quality compounds differently at different scales.
The leader has a specific, honest picture of what isn't working. Coaching isn't a place to go when you're generally interested in getting better. It's a place to go when you can name the thing that's limiting you: the pattern you keep falling into, the conversation you keep avoiding, the way you show up under pressure that you know isn't serving you or your team. The more specific the problem, the more targeted the coaching can be, and the faster the return.
The organization is ready to support the change. This one gets overlooked. A CEO who changes how they lead will need their team to adjust to the new version. If the culture has been built around compensating for the leader's blind spots, coaching the leader without attending to the team simultaneously can create turbulence. A leader who changes in isolation often creates more confusion than clarity.
When it probably isn't
Coaching rarely works when it's mandated. A leader sent to coaching as a performance intervention, or as an alternative to a harder conversation someone doesn't want to have, will likely go through the motions and return unchanged. The research is consistent: commitment and openness to feedback are the primary predictors of coaching outcomes. You can't coach someone into wanting to be coached.
It also rarely works when the goals aren't specific enough to measure. "I want to be a better leader" is not a coaching goal. "My team says I dominate meetings and shut down dissent before it starts, and I want to change that" is. The difference matters because you can't assess whether coaching is working unless you know what working would look like.
And it's rarely worth the investment at the price points most coaching commands when the coach and the leader are a poor fit. Buyers of coaching often optimize for the coach's reputation rather than the match. A coach who is brilliant with one leadership profile can be useless with another. The right question to ask before signing anything isn't "is this person a good coach?" It's "is this person the right coach for the specific thing I need to work on right now?"
The real case for it
The most compelling argument for executive coaching isn't the ROI numbers. It's the access problem. A CEO has almost nobody in their professional life who will tell them the truth about how they're showing up. Their board has agendas. Their team manages up. Their peers are competitors. Their advisors are vendors. A good coach is, structurally, the only person in that ecosystem whose only job is the leader's development, with no agenda other than helping them see clearly and perform better.
If you're running a company or leading a significant part of one, that access is worth a lot. The question is only whether the coach is good enough and the conditions are right. Both of those are things you can assess before you spend a dollar.