70-20-10 learning modelIf Americans called a rose a triantafyllo, the Greek word for the flower, it would still have thorns and it would smell as sweet — or at least, that’s what Shakespeare said, more or less.

While the essence of a rose will remain the same no matter what we call it, the name we give something bares much more significance than we think. People are biased toward names, and there are plenty of weird connections between a person’s birth name and the job they will hold, their athletic performance, their chances of getting an interview, or generally how successful they’ll be in life.

But it feels weird to refer to a rose as a triantafyllo, no offense to anyone Greek. My point is, we should all think twice when we have the power to name something, whether we’re talking about a baby, or a different kind of baby — like a new company, a new invention or a new theoretical concept.

Morgan McCall and his colleagues, who are usually credited with originating the 70-20-10 ratio in the 70-20-10 model for learning and development, probably didn’t get the memo. Or they got it, but they were convinced that including perfectly even numbers was the best way to go. We’ll probably never know. But we shouldn’t hold grudges against them either. The 70-20-10 model is great for organizations.

I’m not against this model of workplace learning. It can be successful if consciously implemented, and my belief is based on science. According to the In-Focus report from Towards Maturity “70+20+10=100: The Evidence Behind the Numbers,” companies that embrace this model are:

  • Twice as likely to analyze problems before designing a solution.
  • Four times more likely to engage directly with managers.
  • Eight times more likely to encourage staff to collaborate online to solve their own problems.
  • Three times more likely to get end users involved in designing solutions.

The report also said that even companies that don’t realize they’re using it witness its positive results.

But 70-20-10 may not be the best name for this particular learning strategy because the numbers are rarely that absolute; they’re guidelines. But the numbers are in the name, so they still raise some eyebrows. For clarity:

  • Seventy percent is for experiential learning and on-the-job discoveries.
  • Twenty percent is for social learning because organizations of all sizes are made of people and people talk with each other.
  • Ten percent is for formal learning with structured instruction, online courses and individual effort.

But why? For instance:

  • Why 10? Ten is an awfully low number compared to the total of 100. Does 10 mean formal learning is not that important?
  • Why 20? People talk a lot. When they’re at work they talk about work as well as personal issues. Sometimes they learn new things, other times they just talk. Should all words uttered in the workplace have a daily cap?
  • Why 70? There are plenty of days when employees must do tasks they already know how to perform perfectly. There are industries where main things don’t change for years. Learning through on-the-job experiences is effective and important, but is it truly 70 percent important?

Again, these numbers are guidelines, so they should depend on each organization’s learning needs. Thus, they can change: 10 may become 19, 20 may become 36, while 70 may become 47. You can make any combination.

The numbers can change and they should change. But why stick to them in the name of this great model if they do change? What should we call 70-20-10 instead? Well, there’s no use reinventing the wheel. Everyone knows what the numbers stand for:

  • Seventy – experiential learning, or experience.
  • Twenty – social learning, or exposure.
  • Ten – formal learning, or education.

Thus, the 70-20-10 model, becomes the three Es learning model, or the Experience-Exposure-Education model. I think the names without the numbers are better.

What do you think? Will we ever drop the numbers?

Livia Mihai is a writer for Matrix by Cypher Learning. Comment below or email editor@CLOmedia.com.



  1. Thoughtful post. Strong words, I suppose, Livia, but I despise the phrase “70/20/10.” I agree with you that we should drop the numbers.

    First, the 70/20/10 name reminds me of the Jack Welch rack & stack method of performance management (20/70/10) where the bottom 10% get lopped off periodically.

    Secondly, and more importantly – to your point in the article – it’s also an inaccurate name and is rarely 70/20/10 in reality.

    Depending on the goals of the organization, especially in my work in guiding sales transformations, the sales support and learning systems that are needed to support behavioral change to drive better organizational results, require much more formal training and support mechanisms than in the average scenario. The 70 and 20 (experience and exposure) leave far too much to chance, if you truly expect to improve business outcomes and lift performance. (See slide #21 in this deck for an example of what I mean: https://www.slideshare.net/MikeKunkle/how-4-sales-systems-can-solve-80-of-your-sales-enablement-challenges)

    Your suggestion of experience, exposure and education is far better, although I’d prefer formal learning or training than “education.” I think of education as acquiring knowledge and influencing mindsets, whereas much of my work is about acquiring and honing skills.

    I’m curious to see other feedback and hope you get some.

  2. I think the 3E’s learning model will be the best name for this. The numbers are truly just a guideline and don’t always turn out to be in that range.

  3. We consciously avoid using 70:20:10 as it has no meaning and can actually cause confusion in a mostly engineering/technically focused corporation. We help focus on Formal, Informal and Social learning (FIS after Dan Pontefract’s model in Flat Army). We develop learning deliverables based on business need using this FIS format to guide our designs. We avoid any form of numerical association as it is not necessary and things change as pointed out in this blog.

  4. There’s no doubt that 3E communicates the essence of the model well. But the names are not only just about what it stands for, but also about getting noticed amidst all the buzz going around. Considering both the aspects, the name has certainly worked: the simple proof is we are still talking about it today even after over two decades 70-20-10 was released. Numbers, I agree, are not absolute, but again the numbers do a lot for stick-ability. A successful predecessor to 70-20-10 that proved it is Mehrabian’s model: 7-38-55, which is almost immortal.

    But coming to TripleE model, what is appealing about it, it is not brushing aside the formal learning events of little significance contributing to to only 10%, but re-terming it gives greater scope for experiential learning can come out of formal training events as well, if the trainers can design ‘meaningful learning experiences’ as part of classroom work.

  5. What I find interesting in this dicussion around numbers, is that we keep looking at them as actual numbers. Arguing whether the 70, 20 or 10 are really true, I find not important. More important to me is the message behind it, and is the 70-20-10, as a concept, helping you to bring that message? Is it helping you to convince people where to put more effort into? If so, it is a good ‘communication vehicle’ and should be seen like that. Common sense will always tell you that the actual numbers will look different. But who actually cares? If it helped you to take the right decisions, and made the difference for your business, it has done its work. A question I always use in this context (when explaining the concept): ‘Can you share with me an important learning or insight you experienced, and at whát moment did it occur?’ You will see, that the majority of people will come up with an example in the 70 or 20, when they tried something out, had success (or failure) or got feedback, that had impact. By asking the question, you’re already getting the message across. And that’s what 70-20-10 is about, to my honest opinion. Far more than the actual numbers…

Add a Comment

Your email address will not be published. Required fields are marked *

Comment *
Name *
Email *