Recently, Training Industry, Inc. released a research report on the 70-20-10 framework. This report seeks to update the concept to the present day from its inception 30 years ago. An important aspect of this research is that we did not seek to refute what 70-20-10 represents with its numbers. But let’s back up a step: What are these numbers supposed to mean?

Based on research conducted in 1987 by Morgan McCall and colleagues at the Center for Creative Leadership, 70-20-10 describes the optimal sources of learning by successful managers. This framework suggests that individuals obtain knowledge, skills and abilities in their jobs through the following mixture of sources:

  • 70 percent from on-the-job experiences
  • 20 percent from social sources, such as interactions with others
  • 10 percent from formal educational events

So, are we saying that these numbers are wrong? Not at all. On the flip side, are we saying that these numbers are right? Not exactly.

Our research showed that there are companies where 70-20-10 is the right mix – but such companies are not the average. Instead, there are a range of mixtures for learning sources.

What does that mean? For some companies, the numbers may be 48-23-29 or 56-27-17 or some other combination. All are valid. But which one is “right?” The answer is that it can change from company to company – and that’s the point. There is no single ratio of learning sources that is best for everyone. That’s why in our research report, we introduced the concept of an OSF ratio.

An OSF ratio represents the relative amount of learning from on-the-job (O), social (S) and formal (F) sources. In this sense, 70-20-10 is simply the OSF ratio that has become popular among L&D professionals. But it would be a mistake to assume it’s the only ratio that works or that it’s the best among many possible ratios. Similarly, it’s a mistake to assume that an L&D function necessarily has to organize its emphasis on learning sources along the 70-20-10 ratio. In fact, we’d recommend against it, if only to force careful consideration of which ratio would better serve business goals.

For some L&D professionals, it may seem like what we’re suggesting is blasphemy. Bear with me, and I’ll walk through a couple examples of why we feel the OSF ratio is a better way to conceptualize sources of learning than adhering to 70-20-10.

Take the example of medical professionals: Do you want your surgeon to be learning most of what he or she knows with a scalpel in hand while patients lie on the table? Or would you prefer the surgeon to have undergone extensive formal training in surgical techniques before ever going into the operating room? Do you want the anesthesiologist putting you under to still be learning what it takes to successfully bring you back to consciousness? In such scenarios, surely there are cases that the surgeon or anesthesiologist will learn from, but these aren’t jobs where formal training on foundational skills can be minimized at the risk of patients.

As another example, take airline pilots: Do you want to get on a plane where the pilot is learning how to take off and land the aircraft while you sit white-knuckled in the cabin?

What about a corporate accountant? Do you want someone learning how to balance the company’s finances and make decisions that are aligned with regulatory and compliance requirements while he or she is on the job? Would it be acceptable to chalk up a financial blunder that could have negative impacts on employees’ continued job security to, “Whoops, I’m still learning how to do this”?

These may seem like convenient examples (and they admittedly are), but they underscore an important point about OSF ratios: An unfortunate side effect of how the 70-20-10 framework is commonly interpreted is that the “O” number is seen as the most critical since it’s the largest number. As shown in the examples above, the “F” number is of the utmost importance – there are jobs where the bulk of learning must take place before an employee ever does anything “in-role” because of the risks of errors. Without the bedrock of formal training, there is no keystone on which to build knowledge and skills through on-the-job learning. While this is true to varying degrees in the above examples, it’s also true to varying degrees for all jobs in any given company.

For some roles, formal training functions as the starting point for deeper learning that will occur later on the job, whereas for others, formal training is a gauntlet that has to be conquered before tackling the job “for real.” In this respect, the “right” OSF ratio can and will vary among companies, among departments within the same company and even among jobs within the same department. So, to base L&D strategy on the assumption that 70-20-10 is right for your company may be making a serious misstep. Instead, learning leaders need to look critically at how formal training plays a role in driving other sources of learning and the quality of that training. Because if it’s ignored, employees will turn to other sources of learning – with no guarantee that what they’re learning will improve performance in any appreciable way.

In our research, we observed that when the training offered by a company is less than effective, employees saw their learning coming from social and on-the-job sources. Note that this case is different than intentionally structuring L&D activities to emphasize these two sources; instead, employees aren’t being served by formal training and are turning elsewhere as a result.

So, why would it be in a company’s best interests to focus on formal training, even if the OSF ratio for that company leans heavily on the “O” slice of the pie? The answer is simple: Formal training is the easiest to control. Learning is an additive game; effective formal training improves the competence of multiple employees, which can make social learning more job-relevant and impactful, in turn making on-the-job learning richer and more meaningful. It essentially becomes a big, tangled feedback loop of knowledge, skills and abilities. To neglect formal training because of a perceived need to have 70 percent of learning occur from job experiences is to willfully short-circuit the loop.

So, for today’s learning leaders, the question is not, “How are we structuring our L&D strategy around 70-20-10?” but, “What is the best OSF ratio for our company?”