Why You Can’t Think Your Way to a Root Cause

You know the feeling in the conference room when the team finally lands it.

The fishbone is covered in sticky notes. You've worked through five whys, or ten, or fifteen. Somebody on the team traces the line with a finger and says, “There it is. That's the root cause.” Heads nod. Pens cap. A warm, settled feeling comes over the room. We figured it out.

That feeling is the most dangerous thing in the building.

The danger isn't that the team is wrong. They might be right. They might even be obviously right. The problem is that what the room just produced isn't knowledge. It's a hypothesis wearing a confident suit.

The Question That Breaks the Spell

The best question I ever learned came from former Toyota people I was lucky enough to work with. They'd ask it plainly, without any edge, right when everyone else was ready to move on. They'd ask it in a lot of situations when something was stated as a fact:

“What do we know, and how do we know it?”

The first time I heard it, I thought it was rhetorical. It isn't. It's a genuine prompt to stop and sort what's been tested from what's been talked through.

When I ask it in a room today, the answer is almost always the same shape.

“Well, we talked it through. We had everyone here. We went to gemba. It makes sense.”

Fair enough. That doesn't make it true.

You Don't Talk Your Way to a Root Cause

This is the move that took me years to fully absorb. A fishbone is not the finish line. Five whys isn't either. An A3 with a beautifully drawn current state isn't either.

They're the starting line.

Thinking and talking get you to a suspected root cause. A hypothesis. Something worth testing. You don't confirm a root cause from a chair. You confirm it by doing something and watching what happens.

The only way to know is to act. Remove what you suspect is causing the problem. Watch. If the problem goes away or gets meaningfully better, you've learned something real. If it comes back when you turn the countermeasure off (when that's safe to do), you've learned more. That's evidence. That's knowledge. That's different from the feeling in the conference room.

And if the countermeasure doesn't work, the reasons aren't obvious from a chair either. It could mean:

• You had the wrong root cause
• You had the right root cause but the wrong countermeasure
• The countermeasure was right but got implemented poorly
• It hasn't had enough time to work

You can't sort those four from the whiteboard. You have to go look. And try. Plan, Do, Study, Adjust. Maybe adjust some more.

The Trap of Feeling Certain

Here's what makes this hard. A good working session feels like knowledge. The sticky notes line up. The logic is tight. Everyone agrees. That feeling is what seduces teams into declaring victory before running the experiment.

It's also what seduces senior leaders into saying things like,

“We know what the problem is, we just need to execute.”

That sentence has sunk more improvement efforts than any technical failure I've ever seen.

If you catch yourself defending an idea instead of wanting to test it, that's the signal. The urge to explain why you're right is a tell that you haven't actually checked.

Small Tests Are Cheap Insurance

The usual pushback is,

“We can't test every idea. We don't have the time or the money.”

Sure. The smallest-test-of-change mindset isn't about being cheap for its own sake. It's about buying insurance.

A few years ago, I was at Mary Greeley Medical Center in Ames, Iowa, and a nurse proposed buying a small caddy with a power outlet for every bed in the hospital. Patients kept dropping their phones and eyeglasses, reaching for them, and falling. A reasonable idea. At about $20 a caddy, outfitting all 220 beds would have run about $4,400.

I asked what the smallest test of change would be.

The nurse said, “One unit.” I asked, “Is that really the smallest?” He smiled. “OK, one bed.”

That's the right answer. Before you spend $4,400, you spend $20 to see if one patient actually uses it. Maybe $100 to try four models in parallel. Think of that $20 as a premium. You're paying up front to avoid spending $4,400 on the wrong thing later.

When the cost of testing is small, the cost of being wrong is small. And when the cost of being wrong is small, people are more willing to actually look at the data. A $4,400 mistake is something you defend. A $20 mistake is something you learn from.

“I Could Be Wrong” Is a Senior Move

There's a perception in a lot of organizations that confidence is the senior move and doubt is the junior one. That saying “we know the root cause” is what a leader does, and saying “how do we know?” is what a nitpicker does.

I'd argue the opposite.

The person in the room who can say “I could be wrong, let's find out” is almost always the most senior thinker there, regardless of title. Certainty is what people reach for when they're worried about looking unsure. Testing is what people reach for when they're confident enough not to need the appearance of certainty.

Taiichi Ohno wrote,

“It is not good if you hold on to your ideas too strongly and try stubbornly to justify them.”

He wasn't being humble for humility's sake. He was describing what actually works over time.

Assuming you're right is more expensive than admitting you could be wrong. Every time.

Back to the Conference Room

So next time you're in that meeting, with the sticky notes on the wall and somebody tracing the line and saying “there it is,” try the Toyota question.

“What do we know, and how do we know it?”

If the honest answer is “we talked it through and it made sense,” you have a hypothesis. That's a good start. The work hasn't happened yet.

What's the smallest, cheapest experiment you could run this week to find out whether you're right?