The 70-20-10 Model: What It Gets Right, What It Gets Wrong, and How to Use It

1. Origins, Evidence, and the Numbers Problem

The 70-20-10 model emerged from research by Morgan McCall, Michael Lombardo, and Ann Morrison published in the 1980s. Their study asked 191 successful executives at American corporations to reflect on what had contributed most to their development. The self-reported distribution, roughly 70% experiential, 20% relationship-based, 10% formal, became the model. The researchers themselves never intended the ratios as a prescriptive formula.

---

2. What the Model Gets Right

The model’s enduring value is as a challenge to training-centric thinking about capability development. Before 70-20-10 became a widely referenced framework, enterprise L&D was almost entirely focused on the formal training event as the development mechanism. The model’s insight — that the developmental experiences most valued by high performers are assignments, feedback, relationships, and stretch challenges rather than training courses — is both intuitively plausible and directionally supported by leadership development research.

1. It correctly names experience as a primary development mechanism. Stretch assignments, new responsibilities, project challenges, and exposure to complexity develop capabilities that formal training can frame and support but cannot produce on its own. An onboarding programme that includes a structured stretch assignment in the first 60 days develops capability faster than one that substitutes additional modules for the same time.
2. It correctly names social learning as underinvested. Peer learning, mentoring, coaching, and expert communities develop capability continuously — yet most L&D investment goes into formal training events. The model usefully challenges the assumption that development requires a designed training intervention. Sometimes the most powerful development occurs when two experienced practitioners work through a problem together and one of them explains their reasoning to the other.
3. It correctly positions formal training as one element, not the whole. The most persistent misuse of formal training is treating it as a complete development solution for a capability that requires practice and experience to develop reliably. The model’s “10” is a useful reminder that formal training is the starting point of a development process, not the completion of it.

---

3. What the Model Gets Wrong

1. The specific numbers are not supported by evidence.  The 70, 20, and 10 ratios emerged from self-reporting by a small, specific population in a specific decade. There is no empirical basis for applying them universally to a manufacturing operative in 2026, a clinical pharmacist, or a software engineer. The percentages create a false precision that leads L&D functions to engineer investment allocations toward these proportions — when the right proportions for any given role and capability depend on the specific learning context, the nature of the capability, and the available development resources.
2. Experiential learning is not automatic. The assumption that 70% of development happens automatically through on-the-job experience is not supported by research. Unstructured experience without feedback, reflection, or appropriate challenge produces habit reinforcement rather than capability development. An employee who performs the same routine tasks for five years does not develop the same capability as one who performs progressively challenging tasks with regular feedback. Experience is a necessary but not sufficient condition for development — the structure around it determines whether it develops capability.
3. It can be used to devalue formal training investment. The model’s “10%” figure for formal training has been used selectively and incorrectly to argue that training programmes are a small portion of development and therefore a small budget priority. This misapplication ignores the model’s own logic: the formal training event that introduces a framework, builds initial skill, and sets the context for experiential practice is the foundation on which the other 90% of development depends. Remove the 10% and the 70% produces undirected experience rather than directed capability development.

---

4. How to Use It Productively Without Focusing on the Numbers

1. Use the model to expand the L&D design brief beyond the training event. For any significant capability development objective, ask: what is the formal learning component (the 10), what structured experiences will develop the capability in practice (the 70), and what feedback, coaching, and peer learning will support the development process (the 20)? This produces a richer development design than a training module alone — without treating the numbers as allocation targets.
2. Structure the experiential component rather than assuming it happens organically. The model’s most practically useful implication is that L&D should design the conditions for effective experiential learning — stretch assignments with defined objectives, manager coaching frameworks, project debriefs with reflection prompts — rather than assuming that time in role produces development automatically. Structured experience produces development. Unstructured time in a role produces repetition.
3. Use the model to challenge training-only responses to capability gaps. When a business requests a training programme for a capability gap, the 70-20-10 lens prompts useful diagnostic questions: could this gap be closed more effectively through a structured assignment than through a module? Is the development barrier a lack of knowledge (training-addressable) or a lack of practice opportunity (experience-addressable)? The model is most useful as a diagnostic prompt rather than a design prescription.

---

In Summary

The 70-20-10 model has earned its longevity in L&D discourse not because the numbers are credible but because the underlying insight is. Most meaningful capability development happens through experience, practice, relationships, and feedback not through formal training events alone. An L&D function that designs only the “10” is investing in the foundation without the structure. One that designs the “70” and “20” as deliberately as the “10” produces more durable capability development.

Using the model productively means taking the insight and discarding the numbers. Ask whether the design of a development programme encompasses experience and social learning alongside formal training — and whether those elements are designed rather than assumed. Do not allocate budgets, headcount, or content investment proportionally to 70, 20, and 10. The ratios have no empirical basis for any specific organisation, role, or capability in 2026.

---

---