WHY GUESSING FAILS

Through both nature and nurture, guessing has become a foundation of our problem-solving skill set. And guessing helps us resolve many of our problems, but only the easy ones. When a light bulb is off, we guess that flipping the switch will turn it on. If that doesn’t work, we guess that changing the bulb will get the job done. If that doesn’t work, we typically scowl at the light as we flip the switch a few more times, and then go check the breaker-box: Aha! We flip the breaker, check the light, and bask in its glow.

What does the IT engineer at your company say when someone calls them to tell them their computer isn’t working? “Is it plugged in?” Often asking three or four such questions solves the problem. If you suddenly start vomiting, you might guess that it has something to do with what you ate last night—and you might be right. But you might not.

Solution-guessing is a hit-or-miss technique. When a problem has two or three potential root causes, and when testing them is cheap and quick, it’s entirely appropriate. But these are easy problems. Most persistent problems in our lives aren’t easy by definition: They would not persist if they were easy to fix.

What would we do if the breaker wasn’t flipped? Or if it flipped again after a few minutes, plunging us once more into darkness? Or if our light bulb blows out repeatedly? At this point, it’s time to realize we don’t have an easy problem on our hands, and guessing won’t solve it. If you don’t have a strong problem-solving skill set, you have three options: You might keep guessing, hoping you might resolve it. You might call in an expert—in this case an electrician—and they’ll be able to use their experience to make an “educated guess,” which can move easy problems along. But when that fails, you’ll probably just cough up the money to replace whatever appears to not be working, or just live with it.

When you’re facing a problem of moderate difficulty, there may be something like 50 potential root causes. Perhaps you’ve developed intermittent sneezing fits, or your motorcycle engine occasionally stalls out in the middle of the highway, or you’re not making any progress on your diet. At work, perhaps your emissions are too close to the regulatory limit for comfort, or you suspect your sales force is not selling as hard as they can because they believe the supply chain won’t be able to meet their commitments

to the customer. If you are really good at guessing—perhaps with the help of some colleagues—you might come up with 30 potential causes.

It takes time and resources to test every guess. With a long list, it’s likely you’ll waste lots of both. Worse, there’s a good chance that the root cause isn’t on your list, and you have no way of knowing until you’ve completed testing the entire thing, which might take months. What will you do next? Perhaps get a bigger group together to create a longer list of guesses?

Then you’ve got hard problems. These are the kinds of problems that might have hundreds or thousands of potential causes. The actual root cause is obscure or hidden. Shearing pipes in your water pipes might be due to invasive corrosive bacteria introduced at the local river. Your trouble sleeping might be caused by an allergy to yellow-6 dye in your macaroni. You are unlikely to be able to guess the causes to these, and trying to guess wastes a lot of time. Trying to implement some of these guesses is a shot in the dark and quickly uses up huge amounts of resources. Your brainstorming efforts will generate a list of some dozens of “possible root causes.” You’ll tirelessly grind through them and, months later, have nothing to show for it. Worse yet, with all of the random changes you have made, you’ve probably created new problems.

Brainstorming might be useful in situations where creativity is required. However, solving hard problems is not one of these. Rather than having one person guess at something, brainstorming is gathering a lot of people together to group-guess, which adds the further complication of groupthink and politics. Often this guessing is covered up with an elaborate “process” for prioritizing the guesses. You can do better than this.

At one food processing plant, they were making a product in a plastic cup with a seal on top—the sort you tear off in order to eat. Customers were getting moldy food because the seals weren’t working properly. You can imagine this was a fairly important problem for brand and food safety reasons. This corporation had invested heavily in Lean and Six Sigma techniques and had a sizeable organization dedicated to solving this problem. When we arrived, they had used a Fishbone-Diagram approach to identify over 200 potential causes and ideas to fix them (this was clearly a pretty hard problem). On the surface, they had taken a very structured approach, but in reality, it’s what I call “structured guessing.” Any time you “come up with” many things to check that could be the cause, you are guessing (see Table 1.1).

If you get from someone a list of 10 “potential” root causes, they don’t know what’s happening. If you’ve come up with 200, you have no idea at all what’s happening. This number of ideas is far too many to search through with any reasonable effort: An individual or team is going to run out of time, resources, and energy long before they get through the list. And worse, when a team doesn’t understand a problem or the system behind it, odds are good that the true root cause isn’t even on the list. This is why guessing won’t solve these problems.

Over a period of 4 months the food processing plant had invested one year of work and $200K trying out about one-third of these ideas, and they’d not gotten close to solving the problem. They had actually created new problems for their production line as they installed new drive-chains for the sealing equipment and made many other changes. When you make 50 changes to a production line, and only one in 10 causes a new problem, you’ve still created five new problems.

Taking an approach designed to solve hard problems took care of this issue in a few weeks, and demonstrated that the root cause wasn’t on the original list. Not a single guess was made in that entire effort. But “structured guessing” had cost the business a lot, including time and money. We’ll have a closer look at this example in Chapter 8, “Make Fact-Based Decisions,” and Chapter 9, “Stay on Target.”