skip to Main Content
Einstein And 70:20:10—L&D’s Odd Couple

In 1905, Albert Einstein had his annus mirabilis, his “year of wonders.” That year, he published four papers that changed the face of modern science. His work covered a range of topics: the photoelectric effect, Brownian motion, mass-energy equivalence, and a radical new view of space, time, and gravity—now known as special relativity.

But ask anyone today to describe a theory of Einstein’s, and most will say “E=mc2,” a shorthand so pervasive it has become synonymous with Einstein himself. But why do we remember this equation over his other—arguably more important—work in physics [1]?

Sticking Power

It’s all to do with the power of stickiness, a concept that applies as much to Learning and Development (L&D) theory as it does to Einstein’s. Sticky ideas are easily retained in peoples’ minds and are easier to recall. Short, simple equations, formulas, and statistics are powerful because they’re often much stickier than the complex ideas behind them.

Stickiness can be both a benefit and a barrier to understanding. E=mc2 refers to the mass-energy equivalence. You can recite the equation, or plug in numbers to do some calculations, but the real value of E=mc2 is what it reveals about our world. For most people, though, it’s an equation without context, and doesn’t have any real meaning beyond high-school physics class.

The danger of Sticky in Learning and Development

In the world of L&D, we find several examples of theories that have been boiled down to inaccurate, but very sticky, L&D mantras: Honey and Mumford’s learning styles, the Myers-Briggs Type Indicator psychometric assessments, and the often quoted Mehrabian myth that only 7% of communication is verbal [2].

Here at GoodPractice, we recently published a report on the 70:20:10 model of workplace learning. This model is one of L&D’s hottest tickets, and is popular both in the UK and around the world. Organizations like Danone, Nike, Dell, and Wal-Mart use it to underpin their learning strategy.

70:20:10 provides a framework for improving and extending traditional training and learning into the workplace. It states that 70% of learning comes from tough on-the-job experiences, 20% from other people, and 10% from formal courses [3].

Though 70:20:10 is popular, our report highlights some major criticisms [4].Practioners of the theory tend to follow a strict interpretation of the ratios. Ideally, the 70:20:10 model should help L&D teams evaluate and optimize their programs. They might find that they’re focusing too heavily on formal learning and not promoting informal and social learning activities. Buy many people in the profession use 70:20:10 as a as a prescriptive set of numbers that can be applied to all learning, regardless of the individual’s or the organization’s unique learning context.

One of our report’s contributors, Nigel Paine, noted:

“The worst elements are people who are determined to get the boxes full up, so they look for the 70%, and they look for the 20%, and they’re not happy until they’ve managed to produce something and say, ‘Well, we’ve now got our 70% of this, and our 20% of that.’ People who over-simplify it get into trouble.”

Is 70:20:10 more than just the sum of its parts

How helpful is it then to have a shorthand for a more sophisticated view? Is 70:20:10 too reductionist a label to be of any proper value? These are good questions and at GoodPractice, and we haven’t settled on a firm viewpoint. In our many conversations with L&D practitioners, we’ve met many people who focus heavily on the numbers and want to use them as a prescription or rule for every learning program they develop. We’ve also seen L&D teams have go beyond the numbers to fully explore the concept.

George Box, a highly respected statistician, famously said:

“All models are wrong, some are useful.[5]

Because all models are a simplification of reality, they can’t be 100% accurate, but they can help us understand, predict, and explain the world around us. If the L&D profession can approach 70:20:10 in a more open-minded way, then the potential for 70:20:10 and its application are greatly increased.

It’s entirely possible that 70:20:10, like other L&D mantras, is doomed to be misinterpreted for years to come [6]. But many L&D experts agree that the 70:20:10 model is here to stay, so we need to propagate a more comprehensive understanding of it. For many people in L&D, the decision to adopt 70:20:10 is an important first step as they work towards a true performance-focused approach to L&D and leave behind the more outdated practices.

Going beyond the numbers

When considering 70:20:10, L&D practitioners can find it challenging to unlock its inherent value and apply this to their own practice. To help L&D teams get the most out of the 70:20:10 model, we’ve identified five key questions.

Five key questions about 70:20:10

  1. How is learning currently viewed in your organization?
  2. Will your organization’s most senior people support and champion 70:20:10?
  3. Does your L&D team have the skills to promote all aspects of 70:20:10?
  4. Is 70:20:10 the right learning mix for your organization?
  5. How can your managers better support learning?

When working with 70:20:10, L&D practitioners can take inspiration from the words of Einstein himself: “Strive not to be a success, but rather to be of value.” [7]

[1] We should note that the mass-energy equivalence is still a vital physical concept, just not as transformational as special and general relativity have been.

[2] The difference here is that, while the public has a surface level appreciation for E=mc2, physicists understand what lies behind the formula. In L&D, there’s often a surface level understanding of our own body of work.

[3] Wikipedia: (accessed 7 December 2014).

[4] Owen Ferguson and Stef Scott, New Perspectives on 70:20:10: A GoodPractice Research Paper. Available at: (November 2014).

[5] Box, George E. P, Norman R. Draper (1987) Empirical Model-Building and Response Surfaces, p. 424, Wiley. ISBN 0471810339.

[6] Mehrabian, Albert (1971). Silent Messages (1st ed.). Belmont, CA: Wadsworth. ISBN 0-534-00910-7.

[7] (accessed 5 November 2014).

Back To Top