Most leadership development programs don’t change behavior. Leadership development is a $50-billion-plus industry, yet in McKinsey’s survey of more than 500 executives, only 11% strongly agreed their programs achieve and sustain the results they wanted. That gap between spend and outcome is the honest verdict after years of watching L&D teams fund workshops, off-sites, and executive retreats that managers enjoy, rate highly on the feedback form, and forget within a month.
The problem isn’t bad content. It’s that a skill taught in a classroom rarely survives contact with a real Tuesday afternoon. This guide is the practitioner’s version of “what works”: which program types actually move the needle, how to pick the right one for your org, and a framework for building something that sticks.
What leadership development programs actually do (and where most fall short)
A leadership development program is a structured effort to build the skills, judgment, and behaviors managers need to lead well, usually through some mix of training, coaching, mentorship, and on-the-job practice. At its best, it turns a strong individual contributor into someone who can set direction, develop others, and make good calls under pressure. That’s the promise.
Now for where it usually falls short. The skills get taught, but they don’t transfer to the job. A manager learns a conflict-resolution model on Thursday, then walks into a tense one-on-one on Monday and defaults to the same avoidance pattern they’ve always used. The knowledge was there. The behavior didn’t change.
This is the gap that defines whether a program works or wastes money. Teaching is easy. Transfer is hard. Everything in this guide points back to that one distinction.
If you want the foundational view of what leadership development is before you design a program around it, start there. This post assumes you already know you need one and focuses on making it actually work.
What works vs. what fails in leadership development
The difference between programs that change behavior and ones that don’t is rarely about content quality. It’s about whether the learning continues after the session ends. The contrast we see most often looks like this:
| What works | What fails |
|---|---|
| Tied to a manager’s real, current challenge | Generic curriculum disconnected from daily work |
| Spaced over weeks with reinforcement | One-and-done workshop or off-site |
| Practice in real situations, then reflection | Role-plays that never resemble reality |
| Measured by behavior change | Measured by attendance and satisfaction scores |
| Built for every level of leader | Budget skewed to the executive tier |
| Individualized to strengths and gaps | One-size-fits-all delivery |
The practitioner verdict is simple. A program fails when learning and doing are separated by time and context. Someone absorbs an idea in a room full of peers, then tries to apply it days later, alone, in a situation that looks nothing like the example. That gap between understanding a concept and using it when it counts is the transfer gap, and it’s where most programs quietly die. McKinsey’s data points the same way: companies with successful programs were four to five times more likely to make managers apply what they learned in real situations and practice it on the job, not just absorb it in a session.
We see this constantly in coaching. The managers who grow fastest aren’t the ones who consume the most material. They’re the ones who get to try a concept in a real moment, then reflect on what happened, while it’s still fresh. Daily practice beats a monthly workshop almost every time.
The 5 types of leadership development programs, compared by effectiveness
There are five common formats, and they differ sharply in how reliably they change behavior. Formal training scales well but transfers poorly. Coaching transfers well but costs more. The honest comparison:
| Program type | Best for | Cost / effort | Behavior-change effectiveness | Biggest failure mode |
|---|---|---|---|---|
| Formal training programs | Emerging leaders, large cohorts, baseline skills | Moderate, scales easily | Low to medium | Content forgotten within 30 days, no follow-up |
| Executive leadership programs | Senior leaders facing strategic complexity | High per head | Medium | Insight without changed daily habits |
| Mentorship programs | Mid-level managers, succession candidates | Low cost, high human effort | Medium to high | Inconsistent mentor quality, fades without structure |
| Coaching programs | Any level needing real behavior change | High (traditional), low (AI) | High | Hard and expensive to scale across a team |
| On-the-job training | Applied skills, stretch readiness | Low cost, high coordination | High when supported | No reflection layer, so lessons go unlearned |
A few honest trade-offs worth naming. Formal training is the default because it’s the easiest to buy and run, not because it works best. It’s fine for teaching frameworks and shared language, but on its own it almost never changes how someone behaves on Monday.
Coaching and on-the-job practice sit at the top for behavior change, and that’s not a coincidence. Both put the manager in a real situation, then add a layer of feedback or reflection. The catch is scale. Traditional one-on-one coaching costs $200 to $500 an hour, which is why most orgs ration it to the executive tier. The strongest real-world programs stack two or three of these types rather than betting on one.
For a deeper look at the structured-course route, see our breakdown of leadership development training courses and which formats hold up.
How to choose the right program, and the right people, for your org
Start with the goal, then match the format to it, then pick the people whose growth will move the organization most. Most teams do this backward. They buy a program first, then look for people to fill the seats. Reverse it.
Match the program type to what you’re actually trying to achieve:
| Your goal | The format that fits |
|---|---|
| Build a shared leadership language across many managers | Formal training |
| Prepare a few senior leaders for bigger strategic roles | Executive program + coaching |
| Develop succession candidates over time | Mentorship + stretch assignments |
| Change a specific behavior (delegation, feedback, conflict) | Coaching, ideally continuous |
| Test readiness before a promotion | On-the-job stretch assignments |
Budget and capacity will stop you from developing everyone at once, so be deliberate about who gets in. Four groups usually deserve priority:
- Skills-gap candidates. Current managers whose specific weaknesses are dragging on team performance. Assess before you decide. A clear baseline tells you who needs support and exactly which skill to target, instead of sending everyone through the same generic track.
- People you plan to promote in the next few months need to start building the manager mindset now, not after the title change. Leadership development and succession planning work best when they run together rather than in sequence.
- High-potential ICs. Strong contributors who show leadership instinct but don’t yet hold the title. Bring in managers to help spot them, then give them real responsibility to grow into. Even those who stay individual contributors become far more effective with these skills. Use high-performer identification to find them rather than relying on gut feel.
- Most programs aim at the top brass. That’s a mistake. Mid-level managers and frontline supervisors touch the most employees daily and have the biggest impact on engagement and retention. Gallup finds managers drive at least 70% of the variance in team engagement, and most of those managers sit in the middle, not the C-suite. Develop them and you change the experience of the whole organization.
The RISE framework for building a program that changes behavior
If you want a program that survives contact with real work, build it on four pillars: Relevance, Individualization, Spaced reinforcement, and Evidence. We call it RISE, and yes, the echo of “Risely” is intentional, because it’s the same philosophy we coach by. Each pillar exists to close the transfer gap from a different angle. It also lines up with what McKinsey’s research found actually drives results: contextualizing the program to the business, designing it for the transfer of learning, and reinforcing change over time, which is exactly what Relevance and Spaced reinforcement are built to do.
Relevance. Managers engage only when the content maps to a problem they actually have. Build the program around real situations from your teams, not textbook scenarios. Before launch, run this test: can each manager point to a moment in the last two weeks where this training would have helped? If not, the content is too abstract, and they’ll mentally check out while staying polite about it.
The second pillar is individualization. No one believes the one-size-fits-all promise, and they’re right not to. A manager drowning in delegation needs something different from one who avoids hard conversations. Assess first, then tailor. The more the program reflects a person’s specific strengths and gaps, the more they’ll trust it and apply it.
Spaced reinforcement is the pillar most programs skip, and it’s the one that closes the transfer gap. Learning that happens once and never returns doesn’t stick. Spacing the same concept across weeks, with a prompt to apply it between sessions, is what moves a skill from “I understand this” to “I did this.” Handouts and follow-up sessions help, but the real lever is a structured nudge to practice in a live situation, then reflect.
Evidence. If you can’t measure behavior change, you can’t defend or improve the program. Define your metrics before launch, tie them to observable behaviors, and use them to determine the ROI of training rather than guessing. Evidence also protects budget. When you can show the program changed how managers act, the next year’s funding stops being a fight.
For a fuller library of structured models you can adapt, our leadership development framework guide and leadership development strategy breakdown both go deeper on the build process. And if you need a personal-level roadmap, leadership development plans covers how individuals turn this into action.
How to measure whether your program is working
Measure behavior change, not completion rates. The most common mistake in evaluation is tracking attendance and satisfaction, then calling the program a success because people showed up and smiled. Completion tells you nothing about whether anyone leads differently now.
Track outcomes at four levels, the classic Kirkpatrick model, and weight the back two:
- Reaction. Did managers find it useful? Easy to collect, weakest signal.
- Did they actually understand the material? Call this learning. A quick check, still not enough.
- Behavior. Are they doing the thing now? This is the leading indicator. If behavior changes, results usually follow.
- Did team-level outcomes move? That’s results. Look at engagement, retention, and performance on that manager’s team, plus promotion readiness across your pipeline.
If you can only pick one, pick behavior. It’s the earliest reliable sign the program is working. For concrete metrics and methods, our guide on leadership development areas and 5 ways to measure them gives you a starting scorecard, and leadership development goals examples helps you set the targets you’ll measure against. If you need to vary how managers practice between checkpoints, these leadership development activities give you options beyond the standard workshop.
A practical tip we give every L&D team: define the behaviors you want to see before the program starts, in plain, observable language. “Communicates better” is unmeasurable. “Gives direct feedback within 48 hours of an issue” is something you can actually watch for.
Why programs fail at transfer, and how daily reinforcement fixes it
Programs fail at transfer because the learning and the moment of need are separated by days, distance, and context. A manager learns a model in a workshop on Thursday. The situation that needs it arrives the following Wednesday, alone, under pressure, looking nothing like the tidy example from class. Without a way to bridge that gap in the moment, the old habit wins.
The fix is reinforcement that lives where the work happens. This is the piece traditional programs structurally can’t deliver, because no human coach can be in the room every Wednesday at the moment a manager needs help. That’s the gap our AI coach, Merlin, was built to close.
Here’s what that looks like in practice. A manager has a tense one-on-one coming up. Instead of digging out workshop notes from three weeks ago, they open Merlin and talk through the actual situation: what they want to say, what they’re worried about, how the other person tends to react. Merlin helps them rehearse the conversation, points out where they’re likely to soften the message, and they walk in with a plan. Afterward, they come back and reflect on what happened. That loop, practice in a real moment followed by reflection, is exactly what closes the transfer gap.
The pattern shows up in the data. Managers using continuous coaching see an average 26% improvement in targeted skills within 12 weeks, and around 73% engage with the daily nudges that prompt them to practice between bigger moments. The point isn’t the numbers. It’s that reinforcement is daily and tied to real situations, which is the one thing a workshop calendar can’t do.
If you’re choosing between vendors and formats, our guide on how to buy a leadership development program walks through the questions that actually predict whether a program will change behavior, not just fill seats.
Where to start
The programs that work share three traits: they’re relevant to real challenges, they reinforce learning daily instead of episodically, and they measure behavior instead of attendance. Everything else is detail. If your current program checks none of these boxes, that’s why it isn’t moving the needle, and it’s also a clear place to start fixing it.
Pick one team. Define the behaviors you want to change. Add a layer of daily, in-the-moment reinforcement and watch whether behavior actually shifts. If you want to see what that continuous layer looks like, try a conversation with Merlin and bring a real situation one of your managers is facing this week.