For decades, organizations have approached leadership development the same way. Identify a group of leaders, put them in a cohort, bring them together for a series of workshops, teach them a few models, hand them a workbook, an action plan and maybe an assessment, then send them back into the business and hope something changes. We have built an entire industry on that model, and it has produced an enormous amount of activity and a disappointing amount of behavior change.
There is only one problem with the model, and it is a fundamental one. Knowing more about leadership does not make someone a better leader. Organizations are beginning to recognize it, and the finance leaders funding these programs recognized it first.
Leadership development is not going extinct. The need for it may be greater than at any point in my career. But the traditional leadership program, as a standalone event with a start date and an end date and a graduation photo, probably is. What replaces it is coaching, and the strongest financial case for coaching is not at the top of the house. It is in the middle.
We Do Not Have a Leadership Training Problem. We Have a Leadership Application Problem.
Companies spend enormous amounts of money teaching leadership. Harvard Business Review has cited more than $46 billion spent annually on leadership training, while pointing to the question every leader faces on the way home from the program: how do you sustain what you learned once you are back in the reality of the business? (Harvard Business Review)
That is the issue, and it has rarely been a content problem. The material in most leadership programs is good. The challenge is what happens Tuesday morning, when a manager is dealing with an underperforming employee, an angry customer, conflicting priorities, an executive demanding results, a talented employee quietly deciding whether to stay, and six meetings before three o’clock. That moment is where leadership development either becomes leadership behavior or evaporates. Most of it evaporates, and the spend evaporates with it.
The data suggests organizations already know this. Gartner reported that leader and manager development remained the number one HR priority for the third consecutive year, while three quarters of managers said they were overwhelmed by the expansion of their responsibilities and 69% of HR leaders said their leaders and managers were not equipped to lead change. More telling still, 75% of organizations had made significant updates to their leadership development programs and were still not getting the results they wanted. (Gartner)
Gartner later reported that 76% of organizations had revised their leadership programs within the previous two years, and nearly 70% of CHROs still said they were not effectively developing senior leaders or midlevel managers. (Gartner)
Read those numbers as an owner rather than as a practitioner. We are spending at scale, redesigning constantly, and the people accountable for the outcome are telling us it is not working. In any other line item, that combination would have triggered a hard conversation years ago. Leadership development has been protected from that conversation because the results have been difficult to see. They are not difficult to see anymore.
The Return Sits in the Middle, and We Have Been Investing at the Top
Here is the uncomfortable economics of the traditional model. When an organization decides an executive matters enough, it gives that person an experienced coach who spends months helping them examine real challenges, think through live decisions, recognize their own patterns, practice difficult conversations and convert insight into action. Then it takes a newly promoted manager with fifteen direct reports and gives that person a two day class.
Something is backwards about that, and it is backwards financially, not just philosophically.
Gallup’s research found that managers account for 70% of the variance in team engagement. (Gallup) Those same frontline and midlevel leaders shape the daily experience of hundreds or thousands of employees. They decide whether people receive real feedback, whether conflict gets resolved or buried, whether strategy gets translated or distorted, whether accountability actually exists, and whether your best people stay.
The cost of getting that wrong is not soft. Gallup put the cost of replacing a single employee at one half to two times that person’s annual salary and estimated that voluntary turnover costs U.S. businesses a trillion dollars a year, with 52% of exiting employees saying their manager or organization could have done something to keep them. (Gallup)
Run that math against your own organization. Take the number of midlevel leaders you have, multiply by the number of people each of them influences, and then apply even a modest improvement in retention, performance or speed of decision making. The leverage in the middle is not close to the leverage at the top. It is an order of magnitude larger, and it is the layer we have historically developed with the cheapest intervention available.
The Evidence for Coaching Keeps Getting Stronger
This is not a philosophical preference. It is where the research has landed.
A 2023 meta analysis published in Frontiers in Psychology examined workplace coaching research and concluded that coaching produces a meaningful positive effect across organizational outcomes, and found no significant disadvantage for virtual coaching compared with face to face coaching. (Frontiers in Psychology) That second finding matters more than it appears, because delivery cost has always been the reason coaching stayed rationed to the executive floor.
A separate meta analysis of randomized controlled studies found significant positive effects on goal attainment, self efficacy, psychological capital and resilience, with particularly strong support for coaching as an intervention that produces behavioral change. (Frontiers in Psychology) Behavioral change is the outcome the traditional program has never reliably delivered.
The International Coaching Federation reports that 72% of respondents associate coaching with increased employee engagement, with strong approval from both senior executives and the employees receiving it. (International Coaching Federation)
There is emerging organizational evidence as well. BetterUp’s December 2025 outcome research reports that coached managers showed a 12% increase in core manager effectiveness, a 32% reduction in voluntary manager turnover, and developed 10% more top talent on their teams than comparison populations. (BetterUp) Those numbers come from a coaching provider’s own research and should be read in that context, but the direction is consistent with everything the independent research shows, and the turnover number alone is worth pricing out against Gallup’s replacement cost range.
The question is no longer whether coaching works for leaders. The question is how we make it available to the leaders who move the business, and how we do it at a cost that survives a budget review.
AI Changes the Economics, Not the Relationship
I do not believe AI replaces great human coaches, and the current research gives us good reason not to make that leap.
A 2025 study comparing human coaches with an AI coaching agent found participants rated human coaches higher across insight, working alliance, goal attainment, commitment and trust. The study was small and exploratory, but the finding reinforces what every experienced coach already knows. The relationship is the intervention. (Human Coaches and AI Coaching Agents)
At the same time, research published in Frontiers in Psychology found that participants developed moderately high working alliances with both human and simulated AI coaches, which suggests people are more receptive to AI enabled coaching than many of us assumed. (Frontiers in Psychology) And newer work is examining what I believe becomes the dominant model: hybrid intelligence, combining the availability and cognitive reach of AI with the judgment, challenge and relational depth of a human coach. (Hybrid Intelligence Framework, Drexel)
That is the future I see. Not human coaching or AI coaching, but human coaching plus AI. And the reason it matters commercially is simple. For years, coaching was the most personalized leadership development experience an organization could buy and the most expensive per leader. AI collapses the cost of practice, repetition and reinforcement, which is exactly the part of coaching that consumed the most hours and the most budget. What is left for the human coach is the part only a human can do.
What This Looks Like in Practice
Imagine a midlevel leader who starts with an assessment that identifies the specific behaviors most tied to success in their role. They meet regularly with a real coach who understands the business, challenges their thinking, builds trust and helps them see patterns they cannot see on their own.
Development does not stop between those conversations. Before tomorrow’s difficult performance discussion, the leader rehearses it with an AI coach, gets feedback, and runs it again. Before presenting a recommendation to the executive team, they practice the argument. Before addressing conflict with a peer, they work through three approaches and choose one. After receiving hard feedback, they reflect on what they heard before bringing it to their coach.
The human coach supplies context, judgment, challenge, accountability and connection. AI supplies access, repetition, practice and immediacy. The workplace supplies the laboratory. McKinsey has made a similar argument about capability building, emphasizing that sustainable change requires people to apply skills in real work, receive feedback, reflect and build new habits rather than complete learning modules. (McKinsey)
The old architecture was learn, return to work, forget. The new one is learn, practice, apply, reflect, coach, practice again. That is a different product entirely, and it produces a different return.
At LAK Group, this is the model we build with clients, because we have sat in the operating seats these leaders occupy. We are not lifelong consultants applying a framework from the outside. We have run the businesses, carried the numbers and built the talent systems, and that is why we design development that has to survive contact with a real Tuesday.
Do Not Eliminate Leadership Programs. Change What They Are For.
There is still a place for the program. Organizations need a common language. Leaders need exposure to new thinking. Cohorts build relationships that pay off for years, and shared experience shapes culture in ways nothing else can. Bringing leaders together has real value.
But the program should be the beginning of development, not the whole of it. Teach the concept together, practice it individually, apply it to live work, coach the leader through the application, measure the behavior, reinforce it and repeat. That is the path from leadership education to leadership capability, and it is the only version of this investment that shows up in business results.
The pressure to get there is building. Gartner found that 85% of business leaders expect skills development needs to increase dramatically because of AI and digital trends, and 93% believe leaders must ensure employees have the time and resources to learn continuously. (Gartner) You cannot classroom your way through that. No curriculum will anticipate everything your leaders encounter over the next three years. You have to build leaders who can think, adapt, reflect and learn while they are in it. That is precisely what good coaching produces.
The Question That Actually Matters
Perhaps leadership programs are not going extinct. Perhaps they are evolving. From events to experiences. From content to behavior. From episodic training to continuous development. From coaching a privileged few to coaching the leaders who run the work. From choosing between technology and human connection to deliberately combining both.
The organizations that see this will stop asking how many leaders they can put through a program. They will start asking a far more valuable question: how many leaders can we make measurably better at leading, and what is that worth to the business?
That is the measure that matters. And increasingly, coaching is how we get there.
Let Us Have the Harder Conversation
If you are funding leadership development this year and you cannot point to the behavior it changed, you already have your answer. Bring us your midlevel leader population, your turnover numbers and your program spend, and let us look at what that investment is actually returning.
Start a discovery conversation with LAK Group, or pilot Coaching On Demand with one cohort of midlevel leaders and measure it against the program you run today. We will help you define the success metrics before we start, because the point is not to run coaching. The point is to prove it moved the business.
Learn more at lakgroup.com.