Image in modal.

In an earlier article, "Finding Root Causes: How to Identify and Understand the Core Issues", we looked at 5-Why analysis as a tool to uncover the root cause of a problem. And we saw that 5-Why is simple, but remarkably powerful.

But in any real-world problem, there are likely to be multiple causes in play at once. So it is helpful to extend your root cause analysis by looking at the problem from several different angles.

The crooked widgets

Let’s say you manufacture and sell widgets, and you get a customer complaint from someone who ordered five widgets that were all crooked. Of course the first thing is to get the customer some good widgets; and right after that, you get a team working to figure out, Why are we making crooked widgets? (And maybe it turns out that the widget-machine is out of alignment because it was left off of the preventive maintenance list.)

But that’s not the whole story. Someone had to collect all those widgets as they came out of the machine, box them up, and ship them out to fill the order. And you might have an inspector who normally counts the widgets in each box to make sure customers get as many as they pay for. With all that handling, you would think somebody would have noticed that they were crooked—after all, your customer saw it right away. So now you have a second question: Why didn’t we detect the problem right away and fix it internally? How did our crooked widgets get all the way out the door and to the customer?

There could be other questions too. Is this widget-machine the only model that’s available, or are there machines that don’t drift out of alignment? And as long as we’ve got this model, what happened with our preventive maintenance program that we missed putting it on the list? In other words, one way or another, How did we get here in the first place?

What is a 3x5-Why?

These are all good questions, and each one shines a light on a different aspect of the problem. We can use 5-Why analysis to drill into each one. And at the end of each path, we’ll find an answer that we can legitimately call a “root cause.” But they are all different. To get a fully-rounded picture we should consider all three.

This is called a 3x5-Why analysis. The point is that—as we saw in the story above—there are three different kinds of root causes, depending on how you approach the issue. And this, in turn, is because there are three ways to think about correcting a problem:

  1. At the moment that a problem occurs, fix it.
  2. If the problem has already happened, catch it.
  3. If the problem hasn’t happened yet, anticipate it and prevent it.

 

A thorough investigation takes all these perspectives into account. In the earlier article, we talked about the first perspective. Now let’s consider the other two.

Why did we not detect the problem?

To address the second perspective, we need to ask a second question (and then drive to an answer using the 5-Why methodology). Besides asking “Why did the problem happen?” we need to ask, “Why didn’t we catch it in time?”

The point is that a working quality system is built on the premise that things go wrong: machines break down, people make mistakes, and so on. As a result, you want your quality system to prevent mistakes before they can happen, and also to catch mistakes after they do. And if a batch of crooked widgets slipped through anyway, there must have been several points of failure.

  • Maybe your line is set up so that the inspectors have no way to report a problem back to the people operating the machine.
  • Maybe your floor personnel haven’t been trained on what a good widget looks like.
  • Maybe the machine automatically packs each unit into a box after building it, so the inspectors can’t see the finished widgets in the first place.
  • Or maybe you offered a huge bonus for getting more units out the door faster, but forgot to explain that shipping bad units doesn’t count towards the quota.

Whatever the reason, finding it and fixing it gives you a second line of defense against the risk of anything going wrong.

Why did we not anticipate the problem?

The third perspective is a little different, because it’s not tied so directly to the exact chain of events that led to the failure. I once had a colleague who explained it by saying, “If you look at it right, everything that ever goes wrong is the fault of Senior Management.” His point was that of course people make mistakes at all levels in the organization. But it is the job of the Quality system to anticipate those mistakes and prevent them. (And Senior Management is responsible to set up the quality system, or see that someone else does.)

  • Someone could make a mistake by not knowing the job. So the quality system includes a training program.
  • Someone could make a mistake by not having the right tools. So the quality system requires an analysis to determine the tools for each task.
  • A piece of equipment could wear out, or could be set wrong. So the quality system includes preventive maintenance and calibration.
  • And so on.

So in the case of the crooked widgets, you want to ask, How did this take us by surprise? Why weren’t we already expecting something like this to happen, and ready for it? In the first place, we already said that the widget-machine was left off the preventive maintenance list, so you have to ask how that happened. More generally, how does the preventive maintenance program keep current with all the equipment it has to service? (Maybe there are more machines missing from the list besides this one!)

But after that, you might start investigating other topics as well. Maybe you’ll review how you set up production in general. For example, you could require a Process FMEA before turning on any new manufacturing line, precisely to anticipate all the ways the line can fail and build in preventive measures, alarms, or contingencies. Or maybe you could look into how you design your products in the first place: sometimes simple design changes can improve overall manufacturability by a lot.

What if we depend on a supplier?

Often the picture is even more complicated, because there is a supplier involved. What if the supplier is responsible for the mistake? In that case, some companies simply push the whole problem-solving effort back to the supplier and wash their hands of it. You can do that, of course, but think twice first. If your customers received faulty product, they got it from you and not your supplier. Shouldn’t you have caught the flaw before passing it along?

In cases involving a supplier, the most complete way to analyze a problem is called a 5x5-Why. But the basic elements are all pieces we have seen before. A 5x5-Why breaks into two parts.

  1. You ask your supplier to carry out a 3x5-Why, as described above.
  2. Then you carry out two branches of the same analysis for yourself. Skip “Why did it happen?” because your supplier is handling that part. But investigate “Why didn’t we catch it?” and “Why didn’t we anticipate it?” with respect to your own operations, independent of anything your supplier says.

The idea is that you already knew your supplier was fallible. So your quality system should have been prepared.

One last word: Digging this deeply can be a lot of work. In general, you should keep your level of effort proportional to the benefit you get from it. So digging this deeply is usually not a good idea for trivial or superficial problems. Keep your focus pragmatic. But when the problem is really serious—and that includes all cases where you have to make a thorough investigation for legal or contractual reasons—these tools are very powerful.