top of page
  • keith

70-20-10 and IT training

chocolate Easter bunnies in school
"But does it work that way in the real world?"

Recently, I was lucky to see a presentation by Paul Gerrard on the skills needed for software testing today. Along with the usual subjects we have come to know and possibly love (planning, test techniques, test management and automation to name a very few), Paul suggested there are a range of "other skills" that testers require.

Along with those comes the challenge of how we can cover both the hard skills like test planning and test techniques along with the other skills suggested by Paul (assertiveness and critical thinking to name two). Of course, those versed in the 70-20-10 model can probably see where this is heading, but for those not familiar with the concept, let me briefly explain.

As the ages have passed, the need for education has changed. In ancient times, learning occurred informally, and usually one-to-one (i.e. "Do what your parent tells you!") As time passed, and individual interactions in a society began to grow, a different kind of learning became more prominent. This could be likened to the ancient Greek symposium - discussion and feedback from peers. Much later, in the 1800's, education evolved to become more production-line in the Industrial Age. Then still later came fun things like multiple choice exams. In the 1970's these were distilled into the 70-20-10 learning model:

  • 70% of learning occurs from challenging assignments and on-the- job experiences

  • 20% of learning occurs from our relationships with colleagues and the feedback we receive

  • 10% of learning happens in formal training, such as courses and workshops

So, there is a loosely coupled relationship in the way people learn. Of course, the 70-20-10 model doesn't consider how individuals learn within these contexts. Whether the learner is a solitary or social learner, and their preferred method of learning isn't considered (visual, verbal, aural, logical and physical - a good site to visit for an overview is

This doesn't mean the classroom is dead. That 10% is an important component. What has gone wrong is the focus on the production-line to get people to pass an exam. There is a failing where people are relying on that 10% to "get the t-shirt" without progressing the learning to the other 90%. Or, after passing the exam, doing the 90% in a haphazard, unstructured way. Many times, I have ranted based on what the syllabus says, knowing in many organisations that process doesn't, and cannot, happen in the way the syllabus states. Thus, without the 90%, the unfortunate case strikes where people answer the multi-choice questions based on "what happens at work" and fail as a result of not following the syllabus.

There is also the question of economics - if one company offers a 3-day course, and another offers a 4-day version, some argue the 3-day would be better as those people attending will still have 2 days back in the office to catch up.

Another point is the speed at which people learn. I have had many colleagues over the years who are methodical learners. They don't just want to know enough to pass, they need to know everything before moving on. Trying to fit learning into 2.5 days with the exam in the afternoon can frustrate them, as well as reduce their chance of succeeding.

So, what's the balance? The 10% is still needed, for certain. But emphasis should be placed on putting that learning into practice. Great, you can answer a question that tells you to use equivalence partitioning. But can you review a user story to determine if you should use that technique or not? Spending time with a specialist in an area is of great value, and a classroom (or virtual equivalent) is an environment where a large amount of information is passed across. Accordingly, we need the time for students to consider this new information, try it out, and discuss this with peers.

Breaking learning into components, with an opportunity to try the learning in the real world to "fail safely" should be the way to go. Developing quality risk managers rather than testers. Just as education changed to suit the needs of the Industrial Revolution, we need to evolve education to suit the Information Age.

Paul is continuing development of the fascinating programme to address some of these issues. The link to Paul's site is here. His work on this subject will be released at the end of November in The Future of Software Quality Assurance. Along with another chapter written by a bloke who rants about security... 😊

22 views0 comments

Recent Posts

See All


bottom of page