Why 70-20-10 model for learning and development is pure unadulterated bullshit

The so called 70-20-10 model (70% workplace or experiental learning, 20% social or informal learning, and 10% structured or formal learning) with which employees worldwide are being brainwashed by their corporate HR comes from a 1986 "research" conducted by Morgan McCall, Michael M. Lombardo, and Ann Morrison at the Center for Creative Leadership (CCL).

The findings were published in the book The Lessons of Experience: How Successful Executives Develop on the Job (1988). Here is how they got to those numbers and why I put "research" under quotes:

• The Sample Size: They interviewed only 191 executives from six major corporations.
• The Question: They asked these executives to look back over their careers and identify the "key events" that made them successful.
• The Results: From these subjective memories, the researchers categorized the responses into:
  • 70% Assignments: Handling tough jobs or "stretch assignments".
  • 20% People: Learning from bosses, mentors, or colleagues.
  • 10% Hardship/Coursework: Formal training and difficult personal experiences.

Their "research" was therefore not a controlled study and it was never published in a peer-reviewed scientific or psychological journal.

If you don't believe me, here is a quote from people who collaborated on it:

The research that led to this publication had its origins in a six-month period (spanning 1981-1982) as a collaboration emerged between four organizations (Union Carbide, Sun Company, Armco and Westinghouse) and three CCL researchers (and eventual authors of the book). The relationship was formalized in CCL’s Research Sponsors Program. Goodyear, PepsiCo and Robinhood Multifoods soon joined the program and the research efforts. The “issue of mutual interest” that bound this group together was how executives learn, grow and change throughout their careers. The data at the core of the book came from 191 successful executives from six major corporation who responded to some form of the question:

When you think about your career as a manager, certain events or episodes probably stand out in your mind—things that led to a lasting change in you as a manager. Please identify at least three key events in your career, things that made a difference in the way you manage now. What happened? What did you learn from it (for better or worse)?

The research established the foundational messages of the book: Executives learn primarily on the job from challenging assignments, hardships and other people. Some experiences are more developmental than others. Different experiences teach different lessons. Yet the book also captures what came to light as the authors worked with human resources professionals and senior line executives to actually implement the finding from the research.

Emphasis mine.

Now that we have established the origin of this absurd model, let's break down the flaws of this "research":

• Survivorship Bias: The data only reflects "successful executives" from a handful of old-guard corporations (like Union Carbide and Goodyear). It doesn't include the people who failed using these methods.
• Post-Hoc Rationalization: The "70-20-10" figures aren't based on real-time observation or measured data; they are based on what managers remembered as important years after the fact.
• Self-reporting: Answers given by managers were taken at face value without questioning either validity or authenticity of data points.
• Zero Peer Review: This was a "collaboration" for a "Research Sponsors Program" intended to help HR professionals "implement findings" not a double-blind study published in a journal of behavioral science.
• Generalization Gap: The research exclusively sampled managers from heavy-industry corporations (Goodyear, Union Carbide, etc.). It completely excluded technical individual contributors, research scientists, and engineers whose successful job execution relies on their formal education.

So why am I writing about this? Because this "model" is being used as a one-size-fits-all for employees regardless of their role.

The 70-20-10 model assumes that all professional growth follows the same trajectory. For a manager back in 1981, "70% experiential learning" might mean learning how to navigate a board meeting by trial and error. However, for an engineer, surgeon, or a pilot, the "70% experiental learning" simply isn't an option -- you don't "learn from your mistakes" while deploying an update to your production database, performing heart surgery or landing a 747 because the cost of experience is a catastrophic failure.

By touting "10% formal knowledge" as a hiring benchmark any HR department following this model is:

• Eroding the knowledge pool not just in their company but everywhere. After all, who will bother to pursue formal education when they are lulled in with lofty promises of "learn 70% on the job"?
• Eroding the employee salaries because those with less formal knowledge have less negotiation leverage.
• Bleeding the talent by putting the strain on a few knowledgable employees to train endless junior hires while doing their job.
• Having a high hire turnover due to low job applicant quality basically wasting resources training new hires for their competitors.

If anything, this model is a perversion of the original 70-20-10 structure where 70% was formal knowledge you brought in, 20% meant informal learning from peers, and 10% of "learning on the job".

Now before you say "but you can't learn everything in school" know that I agree with you. Experience is important and I was a junior once. But what is also important is to understand that junior software engineer should be expected to know the programming language syntax, common algorithms, and to be fluent in use of common developer tools.

Nobody expects junior software engineers to know how a Point-Of-Sale terminal or an ATM work, that's the "learning on the job" part. Teaching them difference between arithmetic and bitwise operators, how to enter BIOS to change settings, how to use a debugger, work with Git, or how to write correct, efficient, and readable code, or just how to find the answer they need using a search engine isn't and should never be part of "on the job learning". It would be like hiring an administrative assistant who not only doesn't know how to use Microsoft Word, but also doesn't know how to read or write.

Conclusion

We need to stop pretending that 191 managers from 1981 found the universal law of learning. We need to hire for competence, respect formal knowledge, and stop expecting our senior talent to be full-time, unpaid faculty for those who decided to take the easy route through life.

A closing word for those who keep propagating this BS model

If you are an HR professional, a "leadership coach," or a manager who continues to propagate this 70-20-10 myth as a universal truth, you need to understand the damage you are doing:

• You are normalizing incompetence. By insisting that 70% of a role should be "learned on the job", you provide a shield for poor hiring practices and low standards.
• You are complicit in age and role discrimination. By ignoring the deep, formal knowledge required for technical roles and focusing on the retrospective memories of 1980s executives, you create a workplace that is hostile to specialists who value objective mastery over subjective "teamwork" in a team where 20% of people do 80% of the work.

TL;DR -- Stop citing a 40-year-old business brochure as if it were a scientific fact because (fact check) -- it isn't. Now get off my lawn!