The 70-20-10 model is a learning and development framework that says roughly 70 percent of what people learn at work comes from challenging experiences, 20 percent from other people, and 10 percent from formal training. It's one of the most quoted ideas in corporate L&D, and one of the most misunderstood.
Most L&D leaders have seen it on a slide. Fewer know where the numbers came from, or that the people who popularized it never meant them as a budget formula. And almost nobody talks about the real gap in the model: experience doesn't turn into capability on its own. Something has to convert it.
This guide covers the origin, the honest critique, how the model works in practice, and what the research on reflection and deliberate practice says the 70 is missing.
Where the 70-20-10 Model Comes From
The model traces back to the Center for Creative Leadership. In the 1980s, CCL researchers Morgan McCall, Michael Lombardo and Ann Morrison asked 191 successful executives to describe the key events that shaped how they managed, and what they learned from each one. The work became the 1988 book The Lessons of Experience: How Successful Executives Develop on the Job.
The executives pointed mostly to hard assignments: turnarounds, big jumps in scope, fix-it jobs, starting something from scratch. Next came other people, especially bosses, good and bad. Coursework showed up, but far less often.
The tidy ratio came later. In 1996, Lombardo and Robert Eichinger wrote in The Career Architect Development Planner that lessons came roughly "70 percent from tough jobs, 20 percent from people (mostly the boss), 10 percent from courses and reading."
Charles Jennings turned that observation into a working framework. A former chief learning officer who went on to co-found the 70:20:10 Institute, Jennings spent years helping organizations use 70:20:10 to move learning out of the classroom and into the flow of work. He's been clear about what it is: "a reference model and not a recipe."
Is the 70-20-10 Model Evidence Based?
Short answer: the direction is supported. The numbers aren't.
The ratio came from executives looking back on their careers and naming what mattered. That's useful data about perception. It isn't a controlled measurement of where learning happens, and it was never tested as a prescription for how to split a development budget.
Researchers have said so plainly. Will Thalheimer flagged the problem back in 2006, noting how rarely real research produces percentages that land on neat multiples of 10. In a 2018 review in Human Resource Development Review, Alan Clardy examined five research traditions behind the claim that most workplace learning is informal. He found the empirical foundation weak, pointing to sloppy scholarship, inconsistent definitions and research protocol problems, and recommended dropping the 70 percent rule. Writing in Public Money & Management in 2021, Richard Harding cited D. Scott DeRue and Christopher Myers' observation that there's no empirical evidence for the ratio, even though scholars and practitioners quote it as fact.
Harding also named the practical risk. A simple ratio makes it easy to justify cutting formal development, on the theory that most learning will happen anyway.
None of this makes the model useless. The core insight holds up: most development happens on the job, and classroom training alone won't build leaders. The mistake is treating 70-20-10 as a measured law instead of a reminder.
How the 70-20-10 Model Works in Practice
Used well, the model is a design lens. Instead of asking "what course should we build," L&D asks "what experiences, relationships and instruction does this capability actually require?" Here's what each part looks like for a first-time manager.
The 70: experiential learning. Stretch assignments, rotations, leading a cross-functional project, owning a turnaround. For the new manager, it's running a first performance review cycle or inheriting an underperforming team.
The 20: social learning. Coaching from a manager, mentoring, peer learning groups, feedback after a key meeting, shadowing a senior leader. For the new manager, it's the debrief with their boss after that first tough review.
The 10: formal learning. Workshops, courses, e-learning, reading. For the new manager, it's the program that gives them a shared language for feedback and coaching before they need it.
CCL's own framing helps here. It describes well-designed coursework as having an amplifier effect on the other 90 percent. The three parts aren't separate buckets to fill. They're supposed to work on each other, which is also the core difference between leadership training and leadership development.
What the Brain Says the 70 Is Missing
Here's the part most 70-20-10 explainers skip. The model assumes that putting someone in a hard job produces learning. Often it doesn't.
Every organization has managers with ten years of experience that's really one year repeated ten times. They've had the assignments. They never extracted the lessons. Two bodies of research explain why.
Reflection turns experience into learning
In 2014, Giada Di Stefano, Francesca Gino, Gary Pisano and Bradley Staats published "Learning by Thinking: How Reflection Aids Performance" as a Harvard Business School working paper. In a field study at Wipro, employees in call center training spent the last 15 minutes of each day writing about the key lessons they'd learned. That time came out of the training day. The control group just kept training.
Trainees who reflected and then explained their lessons to a peer did even better, at 25 percent. The researchers frame it as two kinds of learning running side by side: automatic learning by doing, and deliberate learning by thinking. Their lab studies pointed to the mechanism. Reflection built self-efficacy, people's belief that they could do the task, and that belief drove performance.
Experience isn't the same as practice
K. Anders Ericsson, Ralf Krampe and Clemens Tesch-Römer drew the same line from another direction in their 1993 paper in Psychological Review. Studying violinists, they found what separated the best performers was time spent in deliberate practice: structured activity built specifically to improve performance, with immediate feedback, often under a teacher's supervision, and effortful enough that it isn't inherently enjoyable.
Their conclusion is the one every L&D leader should tape to the wall. Peak performance isn't reached automatically as a function of extended experience, but even highly experienced people can keep improving through deliberate efforts to improve.
Ericsson's team separated practice from work. Work rewards reliable execution. Practice rewards targeted attempts, feedback and adjustment. Most of the 70 is work. A stretch assignment only builds capability when someone helps the leader notice what happened, name the lesson and try again with intent.
The 20 is the converter
That's why the 20 matters more than its number suggests. Coaching, feedback and structured debriefs are how reflection actually happens on a busy Tuesday, and how a repeated task becomes deliberate practice. Notice the Wipro detail: the group that reflected and then shared with a peer did best.
That conversion work is the territory of NeuroCoaching, the methodology Dan Docherty, Braintrust's Chief Coaching Officer, laid out in his book of the same name. The point is managers who can run the debrief, ask the question that surfaces the lesson and give feedback the brain can actually use.
70-20-10 as Written vs. How Most Companies Actually Run It
Plenty of organizations say they run 70-20-10. Here's how the framework's intent usually compares with what happens on the ground.
| Dimension | 70-20-10 as written | How most companies run it |
|---|---|---|
| The ratio | A reference model that reminds you where learning happens | A budget formula, or a reason to cut formal programs |
| The 70 | Challenging assignments chosen for what they'll teach | Whatever the job happens to throw at people |
| Reflection | Implied, but rarely designed in | Left to chance; nobody asks what was learned |
| The 20 | Developmental relationships, especially the boss | An optional mentoring program and managers who were never taught to coach |
| The 10 | Formal learning that amplifies the other 90 | A standalone event judged by attendance and satisfaction scores |
| Measurement | Capability and behavior on the job | Completions and hours logged |
The pattern is consistent. Companies adopt the language of 70-20-10 and keep running the same event-based model, now with a "learn on the job" label attached.
How to Apply the 70-20-10 Model
1. Start with the capability, not the ratio
Pick the specific behavior you need, like a new manager running a hard feedback conversation. Then design backward. Some capabilities need more instruction up front. Others need more reps. The mix should follow the skill, not a fixed split.
2. Design the 70 on purpose
Choose assignments for what they'll teach, not just what they'll get done. Name the learning goal before the assignment starts so the leader knows what to pay attention to.
3. Build reflection into the work
Keep it small and regular. Fifteen minutes at the end of a key week, with a written answer to three questions: what worked, what didn't, and what will I do differently. The Wipro study shows it doesn't take much time to matter.
4. Equip managers to run the 20
Teach managers to coach, debrief and give feedback, then hold them to it. If you only fund one thing, fund this. Our guide to leadership coaching covers what good looks like.
5. Use the 10 to set up the 70
Formal learning should give people a shared language, a clear model and early practice reps, then hand off to on-the-job coaching. A program that ends at the classroom door wastes the amplifier.
6. Measure behavior, not hours
Track whether people do something differently on the job. Completions tell you who showed up. Behavior change tells you whether the 70 converted.
The Bottom Line on 70-20-10
The 70-20-10 model gets the big thing right. Most development happens on the job, and no course replaces a hard assignment. But experience is raw material, not the finished product. It becomes capability when it's paired with reflection, feedback and deliberate practice, and that conversion runs through people: managers who coach, peers who debrief, leaders who ask better questions.
If your 70 isn't converting, the fix probably isn't another program. It's the 20. Worth a conversation? Start a Conversation with our team about what NeuroCoaching looks like for your L&D strategy.
