The Five Whys

If you’re into entrepreneurship and management you’ve probably heard about Lean, Kanban and Kaizen. These techniques were originally developed by Toyota corporation in 1950s as part of larger process called “Toyota Production System”. TPS is the management philosophy and set of practices that organized manufacturing and logistics for

If you’re into entrepreneurship and management you’ve probably heard about Lean, Kanban and Kaizen. These techniques were originally developed by Toyota corporation in 1950s as part of larger process called “Toyota Production System”. TPS is the management philosophy and set of practices that organized manufacturing and logistics for the automobile manufacturer. It was a precursor to widely used and more generic lean manufacturing approach.

The father of TPS, Taiichi Ohno wrote a book called "Toyota Production System: Beyond Large-Scale Production”. The book itself is hard to read because it’s translated from Japanese, but it reveals another way of thinking that I find fascinating. In the second part of the book Taiichi Ohno describes a technique of finding problem solutions that is very simple yet so powerful. “Ask ‘why’ five times about every matter” — that’s what he advises.

How “Five Whys” technique works

This technique is the most easiest analysis technique you can imagine. Unpredicted problems happen all the time and usually they are just indicators of bigger issues that you’re not aware of yet. You can quickly fix the problem and blame the person who let it happen or you can focus on finding the root cause and prevent these types of problems completely. That’s exactly what “Five Whys” is about.

When using this technique you want to get deep into the problem as much as you can. You might find that the origin of it is quite unexpected, as most of the technical problems turn out to be a process or people problem. Here’s how to use it. Let’s say your production server was overloaded for the whole day. Your top priority would be to investigate the incident and fix the load, but you absolutely must research the reason why this happened right after that. You do that by asking why:

  1. Why the server was overloaded? Because there was a 300% database load
  2. Why there was a 300% database load? Because new code was deployed and it contained a mistake
  3. Why there was a mistake? Because there are no performance tests
  4. Why there are no performance tests? Because we don’t have time for them
  5. Why we don’t have time? Because we don’t have enough developers

Distinguish causes from symptoms

There are several important things to note with this technique. First of all you need to make sure you can see the difference between causes and symptoms. What we tend to call problems are often just symptoms of them, the top of the iceberg. For example, if you see a decrease in team’s performance it might be a result of something totally unexpected which would be a real problem. Defining problems as symptoms hides real causes and results in symptomatic solutions that are only temporarily useful.

Pay attention to cause-effect relationship. Effect is defining what happened. Cause is defining why it happened. The causes are decisions we make and actions we take. Sometimes connection is clear but often determining exact relationship might be difficult. Sometimes an effect has multiple causes, in this cause you need to drill down into each one of them.

Blame Processes, Not People

Remember that the purpose of Five Whys is not to blame. but to uncover the root cause. Most of the problems lie in the processes, not people. Remember an example I was showing about? It’s not uncommon to discover that seemingly technical problem turns into process problem like that. It’s tempting to blame the developer who made a mistake but by fixing the process you can significantly reduce the rate of such mistakes.  If you start blaming people you’re really start fixing the symptom and hiding the real problem. This can result into all sorts of problems within the company.

Five Whys technique is a great way to continuously improve processes that have flaws, and we all know that all processes have them. It’s impossible for a company to have perfectly designed interconnected processes. They only way to evolve is to incrementally improve them by finding problems and fixing them. Gradual and never-ending change is the only way to increase efficiency and get better over time.

Limitations

While many managers and teams use Five Whys technicque successfully, there are certain limitations to it. For example, this method is not going to work when cause is unknown. In this case asking “Why?” Is not going to get you any meaningful answer. Another common limitation is that there may be multiple interconnected causes and you won’t be able to have a distinct answer to your “Why?” question.

This method is also very dependent on the skills and personal qualities of the people doing the assessment. If at least one answer is bad or meaningless the whole process stops to make any sense. It’s also not very repeatable — another team may come up with a completely different set of causes and fixes when given the same symptoms and problems.  

Teruyuki Minoura, Toyota’s former managing director of global purchasing commented that Five Whys requires skill and practical training. He emphasised that it’s important to ask questions on spot right after the occurrence of the even, rather then logically trying to come up with causes later. He states that problem-solving is not a logical excerise but something that should happen on-spot and causes should be directly observed.

You just need to get started

The Five Whys technique is just another method to eliminate waste, overburden and inconsistency from your processes, as TPS states. It respects people and teamwork by focusing on improving the process, not fixing people. It makes you base.your management decisions on long-term goals sacrificing short-term losses. It makes you create continuous improvement process where real problems are surfaced and solved incrementally.  

To use this technique right you need to find time to stop and asses the problems you have. This builds culture of getting quality right from the start. This usually causes a lot of skepticism because after all, who’s got time for all this? The reality is that nobody has. But if you use Pareto principle and focus on the most common 20% problems first you’ll gain 80% of the wasted time for your team. You just need to get started.  

tags