What Should Toyota Do Now?

When I last wrote for Bloomberg BusinessWeek about Toyota, there had been two recalls that covered nearly 5 million vehicles, but it seemed to me that the actual defects were small.Don't get me wrong: People had died and others were claiming to be injured, which is as serious as it gets.In my mind, however, those specific problems—and even the subsequent discovery of brake problems in the hybrid Prius—were far from an indictment of Toyota's engineering and manufacturing processes.Journalists were wildly speculating about why Toyota quality had gone downhill, blaming too-rapid growth, cost-cutting, and weaknesses in the philosophy of the Toyota Way. "This is nuts, " I thought.

The week unfolded with new revelations, including complaints by government safety agencies that Toyota (TM) is secretive and criticism that the company had dragged its heels after safety problems were indicated. On reflection (the Japanese call this hansei), I realized that I never studied how Toyota reacts to safety concerns. I had just been guessing, based on experience that I have had with the high integrity of Toyota executives. The question I have since shaped is not what I know or do not know, but how Toyota should respond to what's become a crisis in customer trust.

Any crisis starts with containment, and that is what Toyota is working on now. Once the crisis is contained, though, Toyota should go back to its tried-and-true management principles, which I spelled out in my 2004 book, The Toyota Way, to investigate the root causes and come up with solutions.

What do we now know about the problem? Over 8 million vehicles have been recalled for three problems—an aftermarket floor mat that, if not clipped down properly, can interfere with the gas pedal; a pedal from one supplier that can get "sticky " because a composite material interacts with moisture over time as it wears; and a software glitch on the 2010 Prius that can cause less-than-a-second's hesitation in braking when the antilock braking system is applied.

letting problems accumulate slowly

For each of these problems a specific cause has to be determined—when it occurred, where, and how. There is no evidence that floor mats were assembled incorrectly because they are put on by dealers. There is no evidence that the sticky pedals were assembled wrongly; that issue rather concerns the specific composite material selected for one part. And in the Prius case, the issue lies in the software code, not in how the module was assembled at the plant. So revamping production does not appear to be in order. It seems remarkable that all the recalls occurred within a six-month period. But in fact the cause was at least six months old in each case—five or more years in the case of the pedal design.

Thus, the real principle violated here is to uncover problems immediately and then solve them, one by one. The issues should not have been allowed to accumulate, so in my view the response to the problems (or lack of response) is more serious than the problems themselves. We have all read that this slow response has been characteristic of Toyota for many years. This is a serious allegation. Some say the unintended-acceleration problem is electronic and has yet to be fully solved—which, if true, is a broader problem for all automakers who use so-called drive-by-wire systems.

The point is that I do not know which of these problems is real, or where and why they occurred. (I would venture that journalists do not, either.) Toyota needs to use its own Toyota Business Practices, or TBP, to identify and solve its real problems.

Here's how that process should play out:

Step 1: Clarify the problem vs. the ideal state

The first question always asked in TBP is: What is the problem compared with the ideal state? The ideal state is to strive constantly to completely satisfy customers and provide them safe and quality transportation—zero defects. The problem statement needs to be clear, concise, and actionable. It should not make assumptions about where the problem has occurred or its underlying cause.

The following would be poor problem statements: "Toyota has lost its way in quality" (too vague and general). "Toyota expanded too fast, which led to overworked engineers—and thus quality problems" (makes an assumption about the root cause that has not been proven and cannot be acted upon). "Toyota has evolved a culture of being secretive and hard for government safety organizations to deal with" (draws specific conclusions that if true, should come out of the analysis).

A better problem definition, based on what is actually known and not assumed, might be: "Toyota's reputation for quality and safety has been called into question because of large recalls and there is evidence that Toyota customers have been injured and even died because of defects in their cars."

Step 2: Grasp the actual situation and see the gaps

If customers are complaining about quality and safety, millions of customers have to bring cars in for recalls, and there is even one injury, the ideal is not being satisfied. At this point we need data to understand the scope and extent of the problem. As I mentioned, the number of cars recalled is not so much the issue; the problems are the issue. A single problem is still a single problem, whether it leads to one car or a million cars recalled. The real problem being investigated is unintended acceleration and that should be the major technical focus. Do we have evidence that there is an electronics problem?

In addition to the technical problems, there are reports that Toyota is secretive and uncooperative with local safety organizations. This needs to be thoroughly investigated, as does the ancillary problem of Toyota's public response during the crisis itself. It is critical at this point to be brutally honest. Toyota President Akio Toyoda has committed to bringing in outside quality control experts. This is one of the places they can be most valuable—using unbiased outside eyes to understand the problems.

Step 3: Break down the problem and set targets

The overall problem of quality and safety and responses to customers and governments is too massive to act upon. It needs to be broken down into a smaller set of more tractable problems. At this point sub-teams might be set up to focus on more targeted areas such as development processes, customer-complaint management, engineering changes, public relations, and government relationships. It may or may not be the right time to set specific quantitative targets, but more detailed investigation would be needed.

Step 4: Analyze underlying causes

Toyota is famous for its five-why analysis. Why are there recalls? Because it has been determined that there are serious problems in some vehicles that might be present in other vehicles. Why were there serious problems? Because errors were made in the design stage internally or when working with suppliers? And so on, until the root cause is identified. There would be separate analyses to investigate why it takes so long to respond to customer complaints and why the government believes Toyota is unresponsive to safety concerns.

Steps 5 to 8 would be where Toyota actually comes up with potential solutions (called countermeasures), puts them into practice, monitors the processes and results, and then standardizes and spreads practices that have proven effective.

As you can imagine, this will not happen overnight—and it shouldn't, despite the pressures of the news cycle and Washington politics. Toyota has proven time and again that good planning should be 80% of the effort. Then implementation will go smoothly. It is always a mistake to jump into solutions without thoroughly analyzing the problem and taking time to develop a good plan.

The real test for Toyota will be the longer-term, full problem-solving process. Akio Toyoda seems intent on doing hansei and solving problems the right way, turning over every stone, finding weaknesses, and improving virtually every function of the company. Failure to follow all the principles of the Toyota Way led to this crisis. Now the Toyota Way is the only way out of it.

Before it's here, it's on the Bloomberg Terminal.