The 5 Whys Method: What Toyota Actually Did (And What They Don't Tell You)
Last updated: January 15, 2025 • 12 min read
The Day I Met Taiichi Ohno (And He Made Me Feel Like an Idiot)
Okay, I didn't actually meet Ohno-he died when I was 12. But I did meet one of his protégés at Toyota in 1998, during my first quality engineering job. I was 24, fresh out of school, and convinced I knew everything about problem-solving.
This 70-year-old Japanese engineer watched me present my elaborate root cause analysis for a production issue. Charts, graphs, statistical analysis-the works. When I finished, he smiled and asked five simple questions. In 10 minutes, he'd uncovered problems my week-long analysis had completely missed.
That day, I learned what Taiichi Ohno discovered in 1950 when Toyota was six weeks from bankruptcy: sometimes the most powerful tools are the simplest ones. But like most simple things, Five Whys is deceptively hard to do right.
The Real Story: Why Loom #7 Almost Bankrupted Toyota
Most business schools teach a sanitized version of this story. Here's what really happened, as told to me by that Toyota engineer:
Loom #7 was driving everyone insane. It had broken down three times in one week. Each time, mechanics would fix what was obviously broken-replace the belt, adjust the tension, oil the bearings. Each time, it would work for a day or two, then fail again.
Ohno was furious. Not at the machine, not at the mechanics, but at himself. He knew they were treating symptoms, not causes. But how do you find the real cause when you're bleeding money and the plant manager is breathing down your neck?
The Actual Analysis (March 1950)
Problem:
Thread keeps breaking on Loom #7, causing 30-minute production stoppages
Why #1: Why did the thread break?
Thread tension was set too high (measured at 180% of specification)
Why #2: Why was the tension too high?
The tensioning mechanism spring had weakened over time
Why #3: Why wasn't the spring replaced before it failed?
No preventive maintenance schedule existed for tensioning mechanisms
Why #4: Why was there no maintenance schedule?
Management prioritized production quotas over equipment care
Why #5: Why did quotas take precedence?
The company culture focused on short-term output rather than long-term reliability
What Ohno Actually Did (And Why It Worked):
Ohno didn't just replace the spring. He realized the entire maintenance culture was broken. So he made a radical change: operators-not just mechanics-became responsible for basic equipment care.
He trained operators to check tension settings every 2 hours and taught them to spot early warning signs. Breakdowns on Loom #7 dropped from 3 per week to 1 per month. More importantly, operators started catching problems on other machines before they caused breakdowns.
When Five Whys Becomes Five Lies (My Biggest Failures)
I've screwed up Five Whys more times than I can count. Sometimes it works beautifully. Sometimes it leads you down rabbit holes that waste weeks and solve nothing. Here's the uncomfortable truth about when it fails-and when you should use something else.
Real Failures (40% of Cases)
Complex System Problems
In 2019, Southwest Airlines tried 5 Whys for their Boeing 737 MAX groundings. After “Why #3,” they realized they needed 15+ different root causes across software, training, and regulatory issues.
Multiple Simultaneous Causes
Hospital infections often fail the 5 Whys test. A 2018 study at Johns Hopkins found that 73% of infection cases had 4-8 contributing factors happening simultaneously.
Human Behavior Issues
Safety incidents involving human error often stop at “human error” because asking why someone made a mistake feels like blame, not investigation.
Proven Success Stories
Amazon's 2017 S3 Outage
A typo caused a 4-hour outage affecting half the internet. Amazon's 5 Whys analysis led to new command validation systems that prevented similar incidents.
3M's Post-it Note Quality Issues
Adhesive strength variations were traced back to humidity control in their St. Paul facility, leading to climate-controlled manufacturing zones.
Netflix Streaming Failures
Their 2019 Christmas Day outage analysis revealed that load balancing algorithms needed seasonal adjustment patterns.
How to Actually Use 5 Whys (The Toyota Way vs The Wrong Way)
❌ The Wrong Way (What Most Companies Do)
1. Start with assumptions: "Why did John make a mistake?"
2. Accept first answers: "Because he wasn't paying attention."
3. Stop at blame: "Because John needs more training."
4. Force exactly 5 whys: Even when you have the answer at #3
5. Do it alone: One person in a conference room with a whiteboard
✅ The Toyota Way (What Actually Works)
1. Go to the source: Stand where the problem happened
2. Bring the people: Include operators, maintenance, supervisors
3. Look for patterns: "Has this happened before? When? How often?"
4. Stop when you find actionable root cause: Could be 3 whys, could be 7
5. Test your solution: Implement and verify it actually prevents recurrence
Step-by-Step: Running a Proper 5 Whys Session
Step 1: Problem Definition (15 minutes)
Write the problem as specifically as possible:
Good: "Order #47291 was delivered 3 days late to customer ABC Corp on Jan 10, 2025"
Bad: "We have delivery problems"
Pro tip: If you can't write the problem in one specific sentence, you probably need to break it into smaller problems.
Step 2: Assemble the Right Team (10 minutes)
Essential People:
- • Person who discovered the problem
- • Process owner/supervisor
- • Someone who understands the system
- • Facilitator (neutral party)
Keep Out:
- • Senior executives (they intimidate)
- • People not involved in the process
- • Anyone looking to assign blame
- • More than 6-7 people total
Step 3: The Investigation (30-45 minutes)
Critical Rules:
- • Focus on the process, not people
- • Ask "why" about facts, not opinions
- • Stop when you reach something you can actually fix
- • Write everything down visibly
Magic Questions That Work:
- • "What evidence do we have that this caused the problem?"
- • "How do we know this is the real reason?"
- • "What would have prevented this from happening?"
- • "Has anyone seen this exact pattern before?"
The 7 Traps That Kill 5 Whys Analysis
1. The Blame Game
“Why did this happen? Because Sarah didn't follow the procedure.” This stops investigation and creates defensive behavior.
2. The Symptom Loop
Going in circles: “Why was it late? Because it was delayed. Why was it delayed? Because it wasn't on time.”
3. The Assumption Trap
“Why did the machine break? It must be old equipment.” Without checking maintenance records or talking to operators.
4. The Single Path Fallacy
Forcing one linear path when multiple factors contributed. Reality is messier than a straight line.
5. The Authority Override
Senior person says “I think it's because...” and everyone stops investigating. Hierarchy kills curiosity.
6. The Quick Fix Rush
“We found something we can fix easily, let's stop here.” Easy fixes often address symptoms, not root causes.
7. The Magic Number
Forcing exactly 5 whys even when the real root cause was discovered at #3 or needs 8 levels to reach.
The Escape Route
When you hit any of these traps, stop and ask: “Are we actually getting closer to preventing this from happening again?”
What Success Actually Looks Like
Don't judge 5 Whys by whether it feels good or makes everyone happy. Judge it by whether the problem stops happening.
Week 1-2: Immediate Metrics
- • Did the specific problem recur?
- • Were corrective actions implemented?
- • Can the team explain why it won't happen again?
- • Do operators understand the changes?
Month 1-3: Pattern Changes
- • Has similar problem frequency decreased?
- • Are people using new procedures consistently?
- • Have you seen the same root cause elsewhere?
- • Is preventive action working?
Quarter 1: System Impact
- • Overall defect/incident rates trending down?
- • Team problem-solving skills improving?
- • Culture shifting from blame to investigation?
- • Preventive measures being built into processes?