The 70-20-10 Learning Model: A Complete Guide for L&D Leaders [2026]

The 70-20-10 learning model is a framework that describes how people build skills at work: roughly 70% through hands-on experience, 20% through learning from others, and 10% through formal training. It captures a simple truth that traditional training often ignores, which is that most real capability is built by doing the job, not by sitting in a course.

The value of the model is not the exact numbers. It is the shift in where you put your attention. Instead of pouring the budget into formal courses and hoping they transfer, the model pushes you to design the experience and the coaching that actually creates performance.

Why the 70-20-10 Model Matters Now

Most L&D budgets are inverted. Organizations historically spend the large majority of their training money on the formal 10%, because courses, workshops, and completion rates are easy to buy and track. The model's whole point is that this is backwards: the 90% that drives real capability gets the least investment.

A few pressures make getting this right more urgent in 2026:

• Skills expire faster than courses can be built: With the half-life of many professional skills now measured in a few years, no formal curriculum can keep pace. Experience and just-in-time learning are the only things fast enough.
  

• Work is distributed: The social 20% used to happen at the water cooler. With hybrid and frontline teams, that informal learning has to be intentionally rebuilt through digital coaching, cohorts, and communities.
  

• Leadership wants proof: Reporting course completions no longer satisfy executives. The model pushes L&D to measure the business outcomes that experience and coaching actually move.

What Is the 70-20-10 Learning Model and How Does It Work?

The 70-20-10 model states that effective workplace development comes from three sources in rough proportion: 70% from challenging on-the-job experience, 20% from developmental relationships such as coaching and feedback, and 10% from formal courses and structured training.

For a Head of L&D, the value is not the ratio. It is the design decision it forces. The model moves the question from "what courses should we run?" to "what experiences, coaching, and formal support does this person need to perform?" That reframing changes what you build, how you spend, and what you measure.

Three practical consequences follow:

• You stop treating formal training as the whole program. A course is the launchpad for the other 90%, not the destination.
  

• You design the experience deliberately. Stretch assignments, job rotation, and problem-solving become planned parts of development, not accidents.
  

• You build in coaching and feedback. The social 20% only happens if you create the mentoring, peer review, and reflection that make it real.

The three components map neatly onto the three ways people actually build capability.

• 70% experiential: Stretch assignments, cross-functional projects, problem-solving, and trial and error on real work.
  

• 20% social: Mentoring, peer coaching, feedback, and communities of practice.
  

• 10% formal: Courses, eLearning, certifications, and workshops.

The experiential 70% is closely related to experiential learning theory, which explains why learning by doing, reflected on properly, outperforms passive content.

Learning Model vs the 70-20-10 Model: What's the Difference

These two terms get used loosely, so it is worth separating them.

A learning model is the general category: any framework that explains how people acquire skills and knowledge and gives L&D teams a structure to design around. There are many, each emphasizing a different mechanism. Cognitive models focus on how the brain processes and retains information, experiential models focus on learning through doing and reflection, and blended models focus on mixing online and instructor-led delivery.

The 70-20-10 model is one specific model within that category. Its particular contribution is not how people learn in the brain, but where learning happens and how an organization should distribute its attention and budget across experience, relationships, and formal training.

In short, every 70-20-10 program is built on a learning model, but not every learning model is 70-20-10. The reason 70-20-10 became so widely adopted is that it is less about learning theory and more about resource allocation, which is exactly the decision an L&D leader has to make when planning a program. For a wider view of the frameworks it sits alongside, see our guide to the main types of learning theories.

Where the 70-20-10 Model Came From

The model has a specific origin that explains both its strength and its limits. In the 1980s, researchers Morgan McCall, Michael Lombardo, and Robert Eichinger at the Center for Creative Leadership studied how successful executives said they had learned to do their jobs. Their findings, published in 1996, reported that roughly 70% of development came from challenging assignments, 20% from developmental relationships, and 10% from formal coursework. The Center for Creative Leadership still frames it as a research-based guideline for developing managers.

The framework was later popularized for the digital age by learning experts such as Charles Jennings and Jay Cross, who used it to argue that organizations should stop over-investing in formal courses and instead support learning in the flow of work. Many companies rebranded the three pillars as the "Three Es": Experience, Exposure, and Education.

This history matters because it also reveals the model's main weakness, which the next section addresses honestly.

The Honest Case: Criticisms and How to Use the Model Anyway

Most guides present 70-20-10 as settled fact. It is more useful to be straight about its limits, because that is what tells you how to apply it well.

• The numbers are not scientifically proven: The ratios came from executives self-reporting how they recalled learning, not from a controlled study. Later attempts to replicate the exact percentages produced different numbers depending on the role. Treat the figures as a rough guide, not a law.
  

• It was based on a narrow sample: The original research studied successful senior executives, so the ratios do not automatically transfer to beginners, frontline workers, or high-risk roles.
  

• It fails for novices and high-stakes work: You cannot train a surgeon, pilot, or nuclear operator on a 70% trial-and-error basis. Beginners and regulated roles need a much heavier formal component up front before they can safely learn by doing.
  

• The 70% can become "sink or swim.": Without structured support, "learning by doing" becomes an excuse for handing people hard assignments with no guidance, which drives burnout, not growth.

The practical takeaway is not to abandon the model but to flex it. A software team might run closer to 80-10-10; a manufacturing floor or a clinical setting might need 40-20-40, front-loading formal safety training before any hands-on work. The ratio should follow the risk and the experience level, not the other way around.

70-20-10 Learning Model Examples by Role

The framework looks completely different depending on the job. These examples show how the three components play out across roles, which is the fastest way to understand the model in practice.

1. Corporate leadership development

• 70% experience: Leading a cross-functional restructuring project or turning around an underperforming team.
  
• 20% social: An executive sponsor for regular coaching, plus 360-degree feedback.
  
• 10% formal: A short executive-education workshop on strategy or negotiation.

2. Software engineering

• 70% experience: Owning a real feature or fixing live production bugs during a sprint.
  
• 20% social: Peer code reviews, pair programming, and engineering knowledge-sharing sessions.
  
• 10% formal: A self-paced cloud or architecture certification.

3. B2B sales

• 70% experience: Managing a live territory, running real demos, and drafting client proposals.
  
• 20% social: Shadowing top performers and reviewing recorded calls with a manager.
  
• 10% formal: A product bootcamp or a sales-methodology course.

4. Frontline and customer-facing roles

• 70% experience: Handling live customer interactions and resolving real issues on shift.
  
• 20% social: Escalating complex cases to senior colleagues and learning from the resolution.
  
• 10% formal: Short mobile microlearning on products, service standards, and compliance.

The pattern holds across all of them: the formal layer builds the foundation, the social layer accelerates it, and the experiential layer is where capability actually forms.

Benefits of the 70-20-10 Approach

When development is designed around all three components, the returns show up in metrics a Head of L&D already reports on.

• Faster time-to-productivity: People build real capability on the job rather than in a classroom that may not transfer.
  

• Higher retention of skills: Applying learning immediately, in context, beats passive content that fades to the forgetting curve.
  

• Lower training cost at scale: Leaning on experience and peer learning reduces reliance on expensive external courses.
  

• Stronger engagement: Stretch assignments and autonomy keep people motivated in a way lectures do not.
  

• Better internal mobility: Structured experience and coaching build a pipeline of people ready for bigger roles.

How to Measure the 70-20-10 Model

Because most of the model happens informally, you cannot measure it with completion rates alone. Each component needs its own kind of evidence.

• The 70% (experience): Business and operational metrics such as time-to-productivity, error and rework rates, and performance on stretch assignments.
  

• The 20% (social): Behavioral signals such as 360-feedback shifts, participation in mentoring and communities, and manager-observed change.
  

• The 10% (formal): Application-focused checks, not just scores, such as a 30-day follow-up asking how often a skill has actually been used.

The single test that separates a real program from a training log is whether you can show if capability changed on the job, not just that courses were completed.

How to Choose a Platform for 70-20-10

A traditional LMS was built to manage the formal 10%, which is why so many organizations struggle to support the other 90%. When evaluating platforms, the question is whether the technology can carry all three components. These are the criteria that matter.

• Experiential support: Tools that push guidance into the flow of work and help assign and track stretch projects, not just host courses.
  

• Social learning features: Cohorts, peer feedback, mentoring, and communities of practice, built in rather than bolted on.
  

• Microlearning and just-in-time delivery: Short, mobile content served at the moment of need, not week-long courses.
  

• Skills intelligence: The ability to map learning and experience to roles and skills, so development connects to capability.
  

• Analytics that tie to outcomes: Insight into skill growth and business impact, not just completion dashboards.

Platforms for the 70-20-10 Model: Where Each Fits

Running 70-20-10 manually across a large workforce is not realistic. Scheduling stretch assignments, wiring up coaching, and reinforcing formal learning by hand does not scale. A modern platform becomes the engine that supports all three components in one place.

The table below maps leading enterprise platforms to the 70-20-10 job each is strongest at and the primary use case it serves. It reflects how the market positions these tools today.

Each platform serves a real need. The distinction for a 70-20-10 program is how well a tool supports the experiential and social 90%, not just the formal 10%, and whether it measures capability rather than clicks.

How a capability-first platform supports all three pillars

Disprz is one example of a platform designed to support the whole model rather than only the formal 10%, which makes it a useful illustration of what a 70-20-10-ready platform looks like in practice.

• Supporting the 70%. On-the-job training workflows and skills mapping let managers assign real tasks and record capability built through experience, not just courses completed.
  

• Enabling the 20%. Collaborative learning, feedback, and social features turn peer coaching and knowledge-sharing into a structured part of development rather than something left to chance.
  

• Streamlining the 10%. AI-assisted authoring breaks long content into mobile microlearning delivered at the moment of need, so formal learning launches the other 90% instead of replacing it.
  

• Measuring across all three. Skills analytics connect experience, coaching, and formal learning to business KPIs such as time-to-productivity.

What separates a capability-first platform from a formal-only LMS is measurement across all three pillars. As a data point, ROSHN reported a 15% improvement in business outcomes and a 50% reduction in manual work on a Disprz-powered program, with 91% platform adoption, and Wellness Forever reported new employees becoming productive within 30 days across 400+ stores. These are the kind of on-the-job outcomes the model predicts when experience and coaching are supported rather than left informal.

For enterprises in India, the Middle East, and Southeast Asia, regional evidence carries weight, since global vendors often have limited local reference cases. That is worth weighing alongside the platform criteria above rather than treating any single vendor as the default answer.

Conclusion

The 70-20-10 learning model gives L&D leaders a clear reason to stop over-investing in formal courses and start designing the experience and coaching that actually build capability. The ratios are a flexible guide, not a law, and the real skill is adjusting them to the role and the risk, then supporting all three components deliberately.

The practical challenge is scale. Turning experience, coaching, and formal learning into a consistent, measurable program across a large workforce requires a platform built to support the whole model, not just the formal slice. If your next step is connecting development to business outcomes, that is the right lens for evaluating where your current program falls short and what to change.