Never trust neat percentages? 70-20-10 rule

@J3ro3nJ gave me this hint for a new myth, the 70-20-10 rule. We would only learn 10% from the formal learning situation, 20% through others and 70% through experience and practice. Check this video for some animated explanation:

Doug Lynch busted this myth in a blogpost:

Informal learning: 70-20-10. Does informal learning exist? Folks generally would say yes. But where did the ratio come from and what does it mean? It basically comes from our old Greek friend Archimedes. Then there was an academic, Allen Tough, who in the 1960s used the iceberg as metaphor, and somehow it got popularized as being a formula. Archimedes demonstrated why most of the iceberg is below the surface, and that is where the numbers come from. Think about it: what is the ratio? Is it hours? Compensation? How was it empirically studied? It’s like that old game — telephone. Someone took a conceptual metaphor and then made a misguided inference.

But is this correct? The most often quoted source for this rule is by Lombardo and Eichinger or by Morgan McCall, Robert W. Eichinger and Michael M. Lombardo (e.g. this Princeton-page)

From Wikipedia:

The 70/20/10 Model is a Learning and Development model based on research by Michael M. Lombardo and Robert W. Eichinger for the Center for Creative Leadership. The concept states that development typically begins with realization of a need and motivation to do something about it, and that a blend of different learning approaches “in concert” can provide powerful learning. Lombardo and Eichinger stated that “the odds are that development will be”

This last phrase is quite careful, not really saying it is so, but it could probably be. But this is normal, as scientists are most of the time quite careful in what they state (in my humble opinion, I could be wrong 😉 ).

But there is one thing for sure: this statement is about the education of leaders.

“The 70-20-10 rule emerged from 30 years of CCL’s Lessons of Experience research, which explores how executives learn, grow and change over the course of their careers.
The underlying assumption is that leadership is learned. We believe that today, even more than before, a manager’s ability and willingness to learn from experience is the foundation for leading with impact.” (source)

And this is already a big warning light, because can we transfer this insight to other people? Maybe this is the case for potential leaders, but not for people like you and me?

Don Clarck states about this:

“Some have been calling for 70-20-10 to be the new learning model for across the organization, however, since is a prescriptive remedy for developing managers to senior and executive positions, it does not mean that it is a useful model for developing skills in the daily learning and work flows that takes place within organizations because it is being applied in an entirely different context than what it was designed for.” (source)

My biggest warning light is that I couldn’t find any peer reviewed scientific proof about this rule, for sure no proof for successful transfer. Of course I’m just human and could have been overlooking so every clues towards research is more than welcome!

Still… those very neat percentages… Strange, I never get those when I’m doing statistics?

22 thoughts on “Never trust neat percentages? 70-20-10 rule

  1. Thanks for using my explanation above. However, if anyone thinks that the 70:20:10 framework is about round numbers they clearly haven’t ‘got it’.

    Jay Cross explains it pretty clearly here:

    I’ve also tried to put the (mostly academic and usually business development managers running threatened MBA programmes) naysayers’ minds at rest here.

    I don’t believe anyone in the many organisations that are currently using the 70:20:10 framework think that its about the numbers. It’s simply a practical shorthand that reflects the fact that most adult learning occurs in the context of work and not removed from the workplace.

  2. Thank you very much for your reply and the discussion on the blog of professor Segers. I think we have a same goal but a very different approach. I’m also in favor for change, or rather maybe evolution towards better learning.
    But beliefs and gut feelings deep down have sometimes proven right but at other times very wrong. There are a lot of myths going round in the field of education and I think using guessed figures doesn’t help.
    That the informal and in-service training is important, off course, but this not empirically proven rule describes clear proportions that we can’t tell if they are right or wrong. You can say that the figurs are not important but the message is clear: formal learning is not that important. You maybe not saying this on purpose, but the message in the figures tell this story.
    But what if research would show that the ratio is rather 33-33-33? Or 40-10-50? Would the story change?
    I’m open for empirical data who confirms that the word ‘most’ when discussing informal is correct?
    Secondly, as I wrote this is something that is rooted in the development of leaders, but this rule pops up in other forms of education too without any evidence, e.g. in discussions about blended learning.
    I don’t know if you are familiar with what is called hawthorne effect? (and I know the irony of the name, yes) A lot of people subjective experiencing something as true, doesn’t necessarily means it is…

  3. Pedro, I’m all for solid data where it exists, too.

    Why ’70:20:10′ and not 33-33-33? Well, there is an increasing body of empirical and survey data that indicates organisational learning (i.e. the way adults learn and improve their performance in the workplace) is skewed towards learning in the workplace (‘informal’ and social learning). Again, I point you to Jay Cross’s book and blog here

    These survey data, some of them from large studies – for example the Education Development Center’s 2-year work – all point to informal/workplace learning being >50%, In fact, the figures pivot around 70-90%.

    The sensible inference from this is that the ratio of informal is more likely to be in the upper half than in the lower. I realise it is often a challenge for people who (like me) have spent long careers designing and delivering formal education programmes to accept that our hard work doesn’t necessarily support learning where most of it happens.

    You may say that these studies and others that have resulted in similar pattern don’t constitute ‘evidence’. To my mind we certainly can’t draw the conclusion that the split is exactly 70:20:10 but as I noted in my first comment, I don’t think people who are using the framework (and 70:20:10 is a framework/reference model, not a recipe) build their change programmes around the numbers. They do so in response to the observation that informal and social learning constitutes a significant % of learning in organisations – if not the majority, and they should focus on finding ways to exploit that and build more effective and efficient structures for individual and organisational development in their organisation.

    Incidentally, I do know the Hawthorne Effect. Are you suggesting that the results from all these studies is due to Hawthorne?

  4. Me mentioning Hawthorne-effect in my blogpost was related to your remark that most of the people using this rule have experienced a positive effect.
    I did look at the sources Jay Cross mentions on his website, but correct me if I’m wrong but in his blogpost he only mentions books and reports and the sole peer reviewed paper is in fact an opinion and an second level of a non-peer reviewed research?
    The non-peer reviewed research that is mentioned is also mostly self-reporting and not actually looking at learning effects?
    Don’t get me wrong, I’m not of bad will, but as I asked for scientific evidence I’m still left in the cold I guess?

  5. Since my name was mentioned here, I’ll jump into the conversation. I share Charles’ and your goal of evolving better learning but I do not share your belief that peer-reviewed research is prerequisite to understanding a situation and taking action.

    The academic who acts as if nothing is proven until it is written up in a peer-reviewed journal is as misguided as the business executive who insists that the only measure of profit is that which conforms to generally accepted accounting principles.

    In the case of business, the irony is that things you can’t see have become more valuable than things you can. Eighty percent of the value of the Fortune 500 is intangible.
    When Pacioli invented double-entry bookkeeping to monitor Venetian shipping 500 years ago, intangibles didn’t count for anything. Of course, the stock prices of companies such as Google indicate things have changed in our times. (Almost all of Google’s market value is attributable to intangibles; Wall Street invests in the power of G00gle’s ideas and potential, not its plant and equipment.)

    Yet, business managers still act as if something invisible is worthless because it can’t be seen and sized up. Vestiges of Industrial Age thinking about value live on inside corporate walls. ROI is a useful concept, but it’s not if you leave out the intangibles.
    Measuring intangibles involves making judgment calls, so managers often exclude these factors from their calculations. Accountants are uneasy with judgment calls. These people tote up the numbers for things they can see and list intangibles on the side, as if this keeps their calculations pure. This is nonsense.

    If someone in San Francisco asks me the distance to Los Angeles, and I don’t have a precise answer, I don’t refuse to answer. I say it’s about 400 miles. Or six hours via Interstate 5 in a Lamborghini or two leisurely days down Highway 1 along the coast, with mandatory stops at Point Lobos and Big Sur.

    (By the way, addressing another skepticism in the original post, even if I knew that Los Angeles was precisely 381 miles from San Francisco, I’d say 400 miles. It’s not a lie. I round the number off not to deceive but to simplify. It would be a mistake to doubt me because I quoted an even number. The same is true for the exactitude of 70:20:10. It’s not intended to be precise. An approximate answer is better than no answer.)

    Have I heard of the Hawthorn Effect? Yes, I studied that in business school decades ago. I still metaphorically tweak the lights so those around me know (or are faked into believing) that I care about them.
    You’re heard of the Placebo effect? If it helps someone, I’ll give them a magic sugar pill. If this can highlight the importance of experiential learning, I won’t hesitate to call on it.

    However, suggesting that these are the reason experiential works is nonsense.

    I will suggest that business people know more about how people learn in their organizations than academics do. Do you really think Reuters, Goldman Sachs, Nike, Mars, Maersk, Nokia, PriceWaterhouseCoopers, Ernst & Young, L’Oréal, Adecco, Banner Health, Bank of America, National Australia Bank, Boston Scientific, American Express, Wrigley, Diageo, BAE Systems, ANZ Bank, Irish Life, HP, Freehills, Caterpillar, Barwon Water, CGU, Coles, Sony Ericsson, Standard Chartered, British Telecom, Westfield, Wal-Mart, Parsons Brinkerhoff, Coca-Cola, and others embraced 70:20:10 on a whim?

    Pedro, I don’t buy your analogy that because there are lots of wrong-headed myths about education floating around that we should discount what we see until empirical evidence is around to back us up.

    And as the validity of peer-reviewed papers, look at the hundreds of peer-reviewed studies that purportedly find evidence that the notion of learning styles hold water. (Meta-analysis says they are bunk.) Similarly, here are more than a hundred peer-reviewed journal articles on NLP.
    Learning in organizations is not a science experiment under controlled conditions in the lab. Cause and effect in business is never precise unless it is preceded with the phrase “other things being equal.” Trust me, in the real world, other things are never equal. Reality emerges from the interaction of complex adaptive forces. Stuff happens.

    “Sure, Jay,” you say. “This is logical, but you can’t manage what you can’t measure.”

    Actually, the old can’t-manage-can’t-measure meme is totally wrong. Executives manage unmeasured things all the time. As John Wanamaker, the famous Philadelphia retailer, said, “I know that half my advertising budget does nothing for the business, but I don’t know which half.” All managers make decisions under conditions of uncertainty.
    Decision making involves placing bets on the future. Decisions are more often guided by intuition and judgment than they are by numbers. As the investment prospectus reminds us, “Past success is no guarantee of future performance.”

    A lot of the measurement of learning these days lulls learning leaders into thinking they know what’s going on. In most cases, they need to think again. If 70:20:10 wakes people up to the fact that by and large, people learn their jobs experientially, I think it’s done a good service.

    Sorry to unload on you. You touched a nerve. I cannot abide the idea that the academy should question my observations because they have not been peer-reviewed.

    I’m clearly not a member of the academy, so I won’t close by asking you what is truth?

    All the best.


    1. Thanks for joining the conversation.
      I think it is strange that when I ask for scientific proof I get an answer about how relatively peer reviewed science is?
      One of the first posts I wrote on this blog was about Learning styles and I sure know NLP (sigh).
      I don’t think science is only measuring as I am also a convinced qualitative researcher.
      But, if rules are used and figures are being spread, even if they are approximate, and some people say they are science, I only ask for scientific proof.

      If you read carefully what I’ve written so far I always keep the door open that it may be correct and I sure would like your opinion on the element of transfer that I mentioned.

  6. Pedro, what I wrote about peer-reviewed research findings was in direct response to your initial statement that “My biggest warning light is that I couldn’t find any peer reviewed scientific proof about this rule.”

    I’d enjoy the comfort of solid research findings in this area, too, but in their absence, I’ll go with the data we have from various non-peer reviewed studies and surveys and with what my intuition tells me is the right thing to do.

    You quote Doug Lynch, saying he “beat this myth.” Really, by suggesting that 70:20:10 was ripped off from Aristotle? I challenged Doug to debate the validity of 70:20:10 when we were at a conference in Delhi. Doug refused. Why? He told me since there was no peer-reviewed research, it wasn’t worth talking about.

    Let’s back away from this issue for a moment.

    We both want innovation to flourish. Is the best way to do that by limiting discussion to ideas vetted by faculty members? By raising the red flag when something hasn’t been studied scientifically? Outside of academia, many of us are encouraging as much diversity of opinion as we can accommodate because it leads to the best long term choices.

    I don’t think I’m going to convert you to my way to seeing things nor will you convert me to your views. Let’s agree to disagree for now.

    All the best.


    1. I received also some comments through mail and twitter from academics about this and even from someone guiding a PhD-student working on this theme who couldn’t find any research but often articles saying “it’s been said…”.
      Don’t think we would find many people disagreeing about the importance of informal learning, although also one researcher mentioned internal research of a company hinting a revert importance. As the research isn’t publicly published it doesn’t help, although I do think falsification would be a good way to deal with this. Instead of searching for proof of the rule, let’s look for proof of it not being true?

  7. The null hypothesis.

    But Charles and others don’t claim that 70:20:10 is a universal formula. In some situations, if would be a terrible learning mix.

    There’s also the dilemma of people not appreciating how learning is taking place in their own organizations. I was initially skeptical about the relative importance of informal learning until I reflected on it and talked with some researchers.

    Let’s encourage that PhD student to do some field research.

  8. Absolutely. Lets encourage some formal research into the relative roles of [a] informal/workplace and [b] social learning (the ’70’ and ’20’ in the framework) compared to [c] formal training in terms of impacting individual and organisational performance and productivity. The hypothesis would need to be carefully constructed, but the output could prove very valuable. To date most data comes from inter-organisational surveys.

  9. […] If unsubstantiated through peer-reviewed research, why then is 70:20:10 being adopted in varying degrees by savvy businesses (Coca Cola, British Telecom, American Express, Nike, and many more)? Perhaps the uptake is due to its believability. So much of business operates with intangibles, uncertainty and imprecision – the portion of the iceberg under water. An advertising budget is a good example – one doesn’t know which half of the advertising budget benefits the business and which does nothing for it. The same can be said for the learning budget. A great dialogue between educational scientist and debunker, Pedro De Bruyckere, and the person credited with coining the term ‘e-learning’ and proponent of informal learning, Jay Cross, can be found here. […]

  10. Hi,
    maybe a recent quote from an article by DeRue and Myers in The Oxford Handbook of leadership and organization (2014) can shed some light in this discussion:

    The existing research on experience-based leadership development spans across a wide range of different types of experiences, including informal on-the-job assignments (McCall et al., 1988), coaching and mentoring programs (Ting & Sciscio, 2006), and formal training programs (Burke & Day, 1986). A common assumption in the existing literature is that 70% of leadership development occurs via on-the-job assignments, 20% through working with and learning from other people (e.g., learning from bosses or coworkers), and 10% through formal programs such as training, mentoring or coaching programs (McCall et al., 1988; Robinson & Wick, 1992).

    Despite the popularity of this assumption, there are four fundamental problems with framing developmental experiences in this way. First and foremost, there is actually no empirical evidence supporting this assumption, yet scholars and practitioners frequently quote it as if it is fact. Second, as McCall (2010) appropriately points out, this assumption is misleading because it suggests informal, on-the-job experiences, learning from other people, and formal programs are independent. Yet, these different forms of experience can occur in parallel, and it is possible (and likely optimal) that learning in one form of experience can complement and build on learning in another form of experience. Third, it is inconsistent with the fact that a large portion of organizational investments are directed at formal leadership development programs (O’Leonard, 2010). It is certainly possible that organizations are misguided in their focus on and deployment of these programs (Conger & Toegel, 2003), but we are not ready to condemn formal programs given the lack of empirical evidence. Finally, it is possible that the “70:20:10” assumption leads organizations to prioritize informal, on-the-job experience over all other forms of developmental experiences, which some scholars argue allows leadership development to become a “haphazard process” (Conger, 1993: 46) without sufficient notice to intentionality, accountability and formal evaluation (Day, 2000).

    Kind regards Michelle

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.