Author Profile Picture

Jan Hills

Head Heart + Brain


Read more from Jan Hills

googletag.cmd.push(function() { googletag.display(‘div-gpt-ad-1705321608055-0’); });

The 70/20/10 strategy – what does neuroscience say?


Is the 70/20/10 rule still applicable to modern leadership, or is it time we adapted this model for a new generation of leaders?

The number many companies tell me they are striving for in their learning strategy is 70/20/10.

In case you haven’t come across this or have never really understood it, the theory says managers and leaders become successful and effective roughly based on this mix of experience:

  • 70% from tough jobs
  • 20% from people (mostly the boss)
  • 10% from courses and reading

The original idea was about how high-performing managers learned their stuff.

That study, first identified in research by McCall, Lombardo and Eichinger, has since been verified by other studies. The trio published their research in 1996.

Later the idea was modified by learning professionals to align learning practice with strategy.

The most noted of these was Charles Jennings, who worked at Reuters.

Reference model, not a recipe

In his opinion, 70/20/10 is more of a reference model than a recipe, and that’s a good way of putting it to those who slavishly strive to implement it.

The model, in the context of learning strategy, indicates that effective learning and development comes through experiential and social learning in the workplace (the 70% and 20%) rather than through formal workshops or e-learning programmes (the 10%).

Structured and directed learning, so the theory goes, can be useful but it rarely, if ever, provides the complete answer.

How do we build skills then?

It’s true that we usually build skills and capabilities through experience and practice supported by encouragement and personal motivation.

This is what the concept of 70/20/10 is getting at. Whilst this can be helpful in moving away from a reliance on classroom-based learning, many companies have struggled to give the 70/20 parts of the process the necessary backing, and not a few learning professionals find it hard to know exactly how to help them do this.

This is exacerbated in some HR functions that have a shared service model which assigns regular learning programmes to a minimal level of service and removes too much of the expertise which is needed to support social learning and to help managers with on-the-job learning.

According to a survey in 2014 by Lumesse, the talent software firm, less than 5% of companies achieve the 70/20/10 learning mix.

Further problems

There are other problems with this theory that aren’t widely acknowledged. First, with the pace of change occurring in organisations across many professions, the skills and the mind-set for future success is changing.

There isn’t always the expertise available in a workplace where 90% of the learning is meant to be happening (70% on job and 20% from others, mainly the manager).

We usually build skills and capabilities through experience and practice supported by encouragement and personal motivation.

In fact, it’s more than likely that in such a fluid business environment the way things were historically done is not how they need to be done in the future. So, for learners, the experience on offer from senior players may not be relevant, the skills may be out of date and good role models few and far between.

An analysis of the legal profession

Take, for example, a profession such as the law, which sets great store by junior lawyers learning at the feet (as it were) of the master. Historically this has worked well.

As the demands of clients change to a more commercial focus, however, they don’t just want their legal issues solved, they want them solved in a way that meets their particular business strategy and goals - learning from senior, more experienced legal experts is less and less useful to junior lawyers.

They also need role models who ask probing questions, understand the client's business context and probably the competitive landscape, too. That’s not to say these role models don't exist but they are not as numerous as they need to be, nor are they necessarily able to unbundle how they do the 'commercial' part of the role.

Rather than obsessing on the ratio, the focus should be on how formal learning is actually used.

Some recent research by leadership and learning consultancy DDI supports a view that the 70/20/10 model is in need of revision. They conducted a survey of 13,000 leaders and asked them what experiences had contributed to their expertise, where they had learned the most and how they allocate their learning time.

The results, published in the Global Leadership Forecast 2014/2015 found that rather than 70/20/10, the leaders actually spent 55% (on the job) 25% (learning from others in the work place) and 20% (formal learning) - a doubling of formal learning and a drop in informal experiential learning compared to the prevailing wisdom.

Lumesse, the aforementioned talent software firm, say that 50/26/24 is the average learning mix in most companies right now and, given this new research by DDI and the pace of change, it suggests that this looks to be the right mix.

Rather than obsessing on the ratio, the focus should be on how formal learning is actually used, consistently and over a sustained period, back on the job.

Assessing and measuring this is where the real gains can be made.

This article series contains many articles analysing learning from a neuroscience perspective. Want to read more? The full article series can be seen here.

One Response

  1. I feel that this article has
    I feel this article has missed the fundamental point of the 70:20:10 reference model.

    There is no ‘magic number’ organisations should be striving for in their learning strategy. The point is that we all learn most of what helps us do our jobs well in the ’20’ and ’70’ spheres – as part of our daily workflow.

    The Lumesse and DDI reports are total red herrings.

    The Lumesse author starts from the premise that there is a ‘perfect blend’. A plainly false assumption as the ratios of experiential/social/structured learning that come together as effective ways to build high performance will vary not only between industries and organisations, but also between job roles, specific task contexts, level of initial expertise and probably another dozen or so factors.
    Another problem with the Lumesse data is that the ‘ratio’ question is based on a flawed dichotomy. It asks individuals to self-report the percentage of ‘learning mix’ between [a] on the job; [b] informal; [c] formal. If anyone can offer an explanation of the difference between [a] and [b] I would be interested to hear it. Most informal learning occurs on-the-job, but some on-the-job learning may be formal (for instance directed instruction in an apprentice/master relationship).

    The other question in the Lumesse questionnaire displays a similar lack of understanding of the difference between ‘training’ and ‘learning’ as well as the confusion between ‘on the job learning’,’informal learning’ and ‘formal learning’. If this study were to be presented to anyone supervising an undergraduate research assignment it would have been rejected out-of-hand.

    Equally, the DDI study assumes a ‘right ratio’ based on what appears to be a time-based measure (although this is not clear). DDI’s assumption is that 70:20:10 is the amount of time spent splitting learning between on-the-job, learning through others, and formal. Its brief report then goes on to ‘debunk’ the 70:20:10 model as some limited research carried out by DDI (providers of formal leadership development programmes) has indicated that in its sample set (presumably gathered from participants on DDI programmes) the ideal ratio associated with high quality was 52:27:21. The idea that this ratio could be generalised beyond the sample group is stretching the point. It is meaningless.

    The DDI study does make one sensible point, although it bases this an incorrect assumption: it is that ratios can have the effect of emphasising the separation of different types of learning rather than seeing the different ways in which we all learn as synergistic.

    Anyone who has any experience of using the 70:20:10 reference model will understand the importance of continually looking to leverage the different approaches to help individuals, teams and organisations grow capability and performance. The ‘numbers’ and ratios are not some type of target. Nor are the categories – experience, exposure to others, and formal education – a way to separate learning. They are, in fact, a way to extend our thinking about learning and improving performance as a continuum rather than as a series of courses or programmes. Creating cultures of continuous development.

    Many organisations are using 70:20:10 as a framework to help them support learning as part of the daily workflow. Virtually no organisation or learning professional I have worked with has fretted about whether they are ‘hitting’ a specific ratio.

    if you re-read the Lumesse and DDI reports with those companies business models in mind, you will come to different conclusions than those presented in this article.

Author Profile Picture
Jan Hills


Read more from Jan Hills

Get the latest from TrainingZone.

Elevate your L&D expertise by subscribing to TrainingZone’s newsletter! Get curated insights, premium reports, and event updates from industry leaders.


Thank you!

Thank you! Your subscription has been confirmed. You'll hear from us soon.
Subscribe to TrainingZone's newsletter