I have sat through a lot of leadership development budget reviews. The pattern is always the same. Someone proposes a slate of workshops, a certification, maybe an off-site with a guest speaker. Nobody in the room asks whether that money matches how leaders actually develop. That is the real gap behind what I call the 70-20-10 leadership development blind spot: not that formal training is bad, but that it is almost the only thing companies are willing to design and pay for on purpose.
The 70-20-10 model comes from the Center for Creative Leadership, built on decades of research inside CCL's Lessons of Experience programme and carried forward today by CCL researcher Cindy McCauley. The finding is specific and it has held up for a long time: leaders develop 70 percent from challenging experiences and assignments, 20 percent from developmental relationships, and 10 percent from coursework and training. The exact ratio is widely credited to Morgan McCall, Robert Eichinger and Michael Lombardo, who built on CCL's original research through the 1980s and 90s. CCL still owns, publishes and updates the model today.
The Real 70-20-10 Leadership Development Ratio, and Why It Gets Ignored
Look at that ratio again. Ten percent from courses. Ninety percent from the job itself and the people around a leader while they do it. Walk into most corporate learning and development functions, though, and the budget line reads almost the opposite way. Formal programs, certifications and workshops absorb nearly the whole spend. Stretch assignments happen, but nobody designs them as development. Mentoring happens, but nobody structures it, tracks it or holds anyone accountable for it. The 90 percent that does the heavy lifting gets left to chance while the 10 percent gets a line item, a vendor and a project manager.
70%: Challenging experiences and assignments: The largest share of how leaders develop, according to CCL's research, and the share most budgets do not plan for.
20%: Developmental relationships: Mentoring, coaching and honest peers, usually left to whoever a leader happens to work near.
10%: Coursework and training: The smallest share of real development, and the one that absorbs almost the entire budget.
Why the 10% Gets Funded and the 70% Doesn't
This is not a mystery once you look at it as a design problem rather than a training problem. Coursework is easy to buy. It has a price, a start date, a vendor who sends an invoice and a certificate at the end that looks like proof something happened. A stretch assignment is none of those things. It requires a manager willing to hand a leader real responsibility before that leader feels ready, accept the risk that comes with it, and then coach them through the mistakes that follow. A developmental relationship requires two people to keep showing up for each other over months, not a single afternoon in a conference room. Both are harder to schedule, harder to standardise across a workforce, and much harder to put a number on in a board deck. So they get left out of the plan, and the plan becomes whatever is easiest to procure.
This is also why so much formal training fails to change behaviour once people are back at their desks. I have written before about why conventional leadership training fails to move the needle on its own, and the 70-20-10 leadership development ratio is the reason. A two-day workshop can hand someone a model or a vocabulary. It cannot hand them the judgement that only comes from making a real decision, living with the consequence, and adjusting. Training was never built to carry that weight. The budget just keeps asking it to.
There is also a reporting problem sitting underneath the budget problem. A learning and development team can show a chief financial officer exactly how many people completed a course, how satisfied they were, and how much it cost per head. None of that data exists for a stretch assignment that was handled well without any fanfare, or a mentoring conversation that changed how a leader made a call three months later. What gets measured gets funded, and for decades the only piece of leadership growth anyone bothered to measure was the piece with a completion certificate attached to it. The 70 percent and the 20 percent were always happening. They were simply invisible to the systems that decide where money goes next year.
- Design the 70%: Turn existing stretch work into deliberate development by attaching accountability, sponsorship and a debrief to it.
- Design the 20%: Build mentoring and coaching relationships on purpose, with a cadence and an owner, instead of hoping the right pairing happens by chance.
- Keep the 10% honest: Fund coursework as a small, targeted layer attached to real work, not as the entire development strategy.
Designing the 70%: Turning Real Work Into Real Development
Treating the 70% as something you can design starts with a distinction most companies never make: the difference between a stretch assignment and simply more work. Handing someone a bigger workload is not development. It is a bet that they will absorb pressure without support and either sink or swim, and swimming under those conditions teaches survival, not capability. A stretch assignment that actually develops a leader shares a few features.
- It sits just beyond the leader's current proven capability, not miles beyond it
- It carries real accountability, with a visible outcome the leader owns
- Someone is watching closely enough to intervene before a mistake becomes a disaster
- There is a deliberate debrief afterward, not just a move to the next task
- The organisation treats it as a development decision, not an operational convenience
That last point is the one companies skip most often. A stretch assignment gets handed out because someone is available, not because someone is ready to grow into it. Flip that logic and the same piece of work, a turnaround, a new market entry, a cross-functional program, becomes a deliberate development instrument instead of an accident that happened to teach someone something.
Designing the 20%: Building Developmental Relationships on Purpose
The 20% is the piece most leaders actually remember years later. Ask any senior executive who shaped their thinking and they rarely name a course. They name a boss who told them the truth when it mattered, a mentor who let them fail safely, or a peer who challenged a decision before it became a mistake. The problem is that most companies leave this entirely to luck. A new leader might land with a manager who invests in them, or might not. There is rarely a system behind it.
- Pairing new leaders with a mentor outside their direct reporting line, so the relationship is not compromised by performance reviews
- Setting an explicit cadence for coaching conversations, not an open invitation that never actually happens
- Training managers to give direct, specific feedback instead of vague encouragement
- Rotating leaders through relationships with people who see the business differently, not just people who agree with them
I have seen the companies that get this right, and what they share is not a better mentoring program on paper. It is a small set of managers who treat finding and developing talent as core to their own job, not an extra they get to if there is time. That distinction between a program and a habit is most of the difference between the 20% working and the 20% being a slide in an onboarding deck nobody revisits.
What the 10% Is Actually Good For
Formal training is not worthless. It is not that coursework does nothing, it is that coursework does one specific job and gets asked to do three. A well-built program can give a leader language for a pattern they have already half noticed, a model that organises experience they already have, or a shared vocabulary a whole leadership team can use to talk about the same problem. That is a real contribution. It is just a tenth of the job, not the whole job. Treat it that way and it earns its place in the budget. Treat it as the entire development strategy and it is set up to disappoint everyone who paid for it.
- Audit where the money is actually going: Map current leadership development spend against the 70-20-10 categories, not against course titles. Most budgets will show 80 to 90 percent sitting in the 10% column before any other change is made.
- Name three stretch assignments already happening: Look at the leadership team's actual workload and find the roles or projects that already carry real stretch. Turn them into designed development by adding accountability, a sponsor and a debrief.
- Build a mentoring structure, not a mentoring policy: Assign pairs deliberately, set a cadence, and check whether the conversations are actually happening quarterly. A policy nobody is accountable for is not a system.
- Cut training spend that has no attached practice: A course with no assignment, no manager follow-up and no chance to apply the material immediately is close to a wasted seat. Attach every course to a real piece of work within thirty days.
- Report the ratio, not just the training calendar: Put the 70-20-10 split in front of the people who approve the budget. A visible ratio changes what gets funded next year, because it makes the imbalance impossible to ignore.
The Architecture Question Behind the Ratio
This is where I think about leadership development as an architecture problem rather than a training problem. A building needs a foundation, a load-bearing structure and finishes, and all three matter, but they are not interchangeable and they are not equally weighted. Formal training is the finish work. Stretch assignments and developmental relationships are the foundation and the frame. I have written about what causes leadership development programs to fail, and the common thread is always the same: someone tried to finish a building that had no frame. If you want the structure to hold weight once a leader is under real pressure, the 70% and 20% have to be designed with the same intent as the 10%, or the whole thing looks complete and collapses the first time it is tested.
Getting this right is less about adding a new program and more about changing how existing work gets used. I have laid out how to embed leadership development so it actually sticks, and the short version is that development has to live inside the job, not next to it in a separate calendar. A ratio only works as a design principle if someone owns it. Otherwise it stays a slide from a workshop nobody applies.
A leadership development budget that spends almost everything on the 10% is not a philosophical error. It is a structural one, and structural errors compound. Every year the ratio stays inverted, the organisation gets better at running workshops and no better at producing leaders who can carry real weight. Rebalancing it is not expensive in the way a new course is expensive. It is a decision to design the 70% and the 20% on purpose, the same way you would design anything else you actually wanted to work. That is the same thinking behind CapabilityAI and the Architecture Accelerator programme: build the structure first, then let the coursework do the smaller job it was always meant to do.