The Five Whys

Dear Gemba Coach,

I’m a lean coach and teach A3s in my company. Every lean book mentions “5 why,” and I can see the power of it in theory but I struggle with it in practice. Any tips?

First of all, pat yourself on the back for admitting that you find “5 why?” difficult. I’ve been doing gemba walks two or three times a week for longer than I care to remember and can’t recall more than a dozen instances of truly insightful “5 why?” analyses. I clearly remember the last case, a few months ago, in a company that manufactures industrial equipment. The production team was trying to understand a recurring quality problem and finally hit upon the fact that two core elements of the product were not interacting as expected. A member of the engineering department was part of the problem-solving group, and he accepted to take back the issue to his colleagues, where it finally revealed that an opinion commonly held by all engineers was false: they had the wrong mental model about how their product worked. Consequently, they experimented until they fixed the issue on this specific product and are in the process of checking all others. As you can imagine, several planets have to align for this sort of problem to get resolved. First, production doggedly continued to pursue a quality problem it had previously given up on; second, it got engineering involved (such teamwork is never a given, not in this company or in any other I know); third, they rigorously tested factors until narrowing it down to the one they previously thought could not have an impact; finally, they kept asking “why?” until they concluded that their thinking about the physical interaction of components was wrong. It was a truly magical “lean moment” but, on the other hand, it doesn’t happen that often. Many “5 why” exercises I’ve witnessed, remain superficial, run in circles, or end up with changing the process without having first understood the problem (and thus investing on a solution and later on finding out the problem persists).

Let us apply lean thinking to “5 why”: what is the test method for a proper “5 why?” analysis? Process performance gaps are an indication that you are controlling the wrong set of parameters – if you know what parameters affect process performance you fix the problem quickly, it’s a matter of turning the dial back to its standard value. In many cases however poor process capability stems from the fact that we simply don’t know what the driving parameters are. A successful “5 why?” analysis is one that leads you to discover the proper controlling parameter. In Ohno’s classic example (From his book on the Toyota Production System):

1. Why did the machine stop? There was an overload and the fuse blew.
2. Why was there an overload? The bearing was not sufficiently lubricated.
3. Why was it not lubricated sufficiently? The lubrication pump was not pumping sufficiently.
4. Why was it not pumping sufficiently? The shaft of the pump was worn and rattling.
5. Why was the shaft worn out? There was no strainer attached and metal scrap got in.

We can see him shifting from the demand on the machine to the lubrication of the bearing, to the working of the pump, to wearing of the shaft, to the protection of the shaft: in order to solve the problem, we no longer control the overload of the machine, but the inflow of metal scrap in the pump. This classic example also demonstrates why “5 why?” is so tricky: in the same situation anyone with less intimate knowledge of the workings of the machine could have answered differently and taken the “5 why?” to a completely other conclusion. The test measure of a successful “5 why?” analysis is when it has led you to change your mind about what driving parameter to control in order to deliver process capability.

This, in turn, reflects exactly the kind of deep thinking the “Thinking Production System” is supposed to develop. Most technical problems are multi-cause, multi-impact (anything simpler can be solved simply by looking at it). Toyota-style problem solving is about learning to turn a multi-cause multi-impact problem in a one-by-one cause-effect problem by identifying which cause (of all potential factors) is responsible for the largest impact (of all potential impacts). Developing the knowledge to correctly substitute the right cause-effect model to a multi-cause/multi-impact situation is, in effect, wisdom – and requires hours of practice on narrow technical problems. This kind of wisdom allows you to then start the “5 why?” analysis at the right place. If not, who knows where you’ll end up. The key to a successful “5 why?” is a correct grasp on the current state. Without a narrow, specific description of the current state, asking a series of “why,” is likely to lead you… just about anywhere.

A lot of hard work needs to be done before embarking on the first “why?”: potential factors causing the problem have to be identified and tested one by one:

As Art Smalley described in the paper we wrote together (The Thinking Production System), in his first problem solving in Ohno’s engine plant he had to test over ten factors before hitting on the fact that the machine was producing defects because the lubrication liquid was contaminated by bacteria. When he proudly announced his findings to the supervisor (after arduous work to figure it out), the supervisor simply asked: “why was the liquid contaminated?” Art had the correct factor to grasp the current state. Now he had to figure out the root cause.

In my experience it takes a definite maturity for problem solving groups to actually stay the course and do the cause search properly. In most companies, the group just wants to “fix” the issue, and find the first way they can to make the problem go away. “5 why?” has a completely different aim: developing the technical expertise of the people conducting the analysis by finding the correct root cause. The companies I know who are getting there tend to speak less in terms of general processes and more of narrow, specific technical expertise. They come to see that:

1. The leanness of your processes is a reflection of the technical skill of the people in the process: more skilled people can run a tighter ship, and on the other hand, leaning a process without the necessary skill just leads to a well designed poorly performing process;
2. “Lean” is not a thing in itself but a method to develop towering expertise and teamwork (to take Jeff Liker’s terms in The Toyota Way to Lean Leadership), which, in turn leads to better judgment, leaner processes, and sustainable results.

Many managers I meet are uneasy about this. They like the straight line of fixing a process: they can ask (change this) and see a direct result (or not, more likely). The dotted hand of training people and then seeing results improve, although experience shows that it is more effective time and time again, is harder for many managers: it takes a leap of faith in people’s autonomy and attitude they seem unwilling to take. Yet, we all know the economic value of common trust.

To answer your question it’s hard to have specific advice on how you conduct your “ why?” so remote from the gemba other than reflecting on your expectations of what the “5 why?” analysis is and does. To correctly practice “5 why?” you must first master:

1. Having a clear mental picture of how the technical process should work
2. Being able to identify correctly the point of cause where the technical process is misbehaving
3. Listing the potential factors of this variance and testing them one by one until you can confirm the main cause-effect relationship
4. And then asking why?

“5 why,” cannot be practiced at a generic level – it’s a practice to develop deep, narrow-specific knowledge and makes little sense if divorced from the actual technical process it’s being applied to. The greater the expertise, the most relevant the five whys; the lesser the expertise, the more dubious the five whys. I am not saying you shouldn’t practice “5 why?” at every opportunity, it’s a great way to get people to think further than the immediate work around, but that you should accept that the outcome of the exercise is very linked to the depth of technical understanding of the people carrying it out. In effect, the first step to conducting a “5 why?” analysis is to estimating the expertise level of the people in the group and, if necessary, inviting in more knowledgeable people to steer the discussion.

Like any lean tool, “5 why?” is a practice and can’t be taught other than by repeatedly going through it. It takes about 10,000 hours to master any skill, so don’t let yourself be discourage and keep practicing it at any opportunity, but try to keep the ideal in mind – better understanding cause-effect relationships in technical processes – in order to get the full value of every “why,” you ask.