I often use 5 Whys in workshops.
When I have around fifty participants, I may divide them into five groups of ten, give every group the same problem, and ask them to work through the same exercise. You might expect the flipcharts to look reasonably similar because everyone started from the same place.
They rarely do.
By the first why, I may already see two or three different explanations around the room. More interestingly, even within a single group, one why may produce three, four, or five plausible answers. If the group allows those answers to remain visible, the exercise stops looking like the straight line we usually associate with 5 Whys.
It starts looking more like a mind map.
That observation changed how I use the tool.
The value of 5 Whys is not that the fifth answer magically reveals the root cause. Its value is that it helps us get beneath the first explanation, surface different causal paths, and think more carefully about what deserves investigation.
The fifth why is not the truth.
It is one possible place to look next.
One Why Can Have Several Answers
Consider a familiar workplace problem:
Customer complaints take too long to resolve.
Ask the first why and one person may say, “Because frontline employees wait for supervisor approval.”
Someone else says, “Because customer information is difficult to find.”
Another says, “Because complaints are assigned too late.”
A fourth says, “Because another department takes too long to respond.”
All four may be describing something that actually happens.
This creates an important choice.
We can force the group to select one answer and continue asking why along that single path. Or we can allow several plausible answers to remain visible and investigate them separately.
If we choose only the approval issue, we might end up with a chain like this:
Employees wait for supervisors because their authority is limited. Their authority is limited because managers worry about expensive mistakes. Managers worry because recovery guidelines are unclear.
After five questions, someone may write:
Root cause: unclear recovery guidelines.
That may be useful.
But what happened to the missing information, late assignment, and interdepartmental delay?
They disappeared when we chose one answer to the first why.
The fifth answer may be the end of one causal path. It does not automatically explain the whole problem.
The Straight Line Can Hide the Problem
The familiar 5 Whys diagram is attractive because it is neat:
Problem → Why → Why → Why → Why → Why → Root cause
Real organizational problems are often less cooperative.
A delayed customer complaint may involve several conditions at once. Employees may lack authority. Information may sit in another system. Supervisors may be overloaded because too many routine decisions travel upward. Another department may be measured against a priority that makes this particular request less urgent.
Some of these causes may be independent. Others may reinforce one another.
For example, when employees cannot find information, they ask supervisors more often. As supervisors become overloaded, approvals slow down. As approvals slow down, employees become even more reluctant to act without permission.
Now the problem is no longer one chain.
It is a cluster.
This is why I hesitate when a group reaches the fifth why and announces that it has found the root cause.
Perhaps it has found an important contributor. Perhaps it has uncovered one condition sitting upstream of several others. Perhaps it has simply produced a convincing story.
We do not know yet.
A clean diagram is not proof.
Different People See Different Parts of the Work
The differences among groups are not merely noise.
Where people stand in the work changes what they notice.
A frontline employee may immediately see the missing customer information. A supervisor may notice that too many decisions need approval. Someone from Finance may understand why the approval rule exists. An operations manager may see a broken handoff. HR may notice that employees have never practised handling the situation independently.
They are looking at the same visible result from different positions.
That does not mean every explanation is equally correct. People can make assumptions, defend their department, remember events selectively, or confuse an opinion with evidence.
But disagreement can still be useful.
Instead of immediately asking, “Which group got the right answer?” I prefer to ask:
What did one group notice that the others did not?
That question keeps useful information in the room.
The same applies inside one group. If ten people produce four plausible answers to the first why, I do not automatically see that as a problem to solve. The divergence may be showing us that the situation has more than one important cause.
Do not force consensus merely to make the flipchart easier to read.
Let the Whys Branch
When several answers seem plausible, let them branch for a while.
Suppose our slow customer-resolution problem produces three first-level causes:
Waiting for approval
Missing information
Late assignment
Now follow each one.
Why are employees waiting for approval?
Perhaps the decision limit is very low. Why is the limit low? Maybe managers fear inconsistent decisions.
Why is information missing?
Perhaps account history sits in another system. Why? The systems were built separately and nobody designed the handoff around the actual service moment.
Why are complaints assigned late?
Perhaps the queue is checked only twice a day. Why? Nobody owns live triage.
The purpose is not to create the largest possible mind map.
It is to avoid narrowing the diagnosis before you have enough reason to narrow it.
As branches develop, you may notice that some become weak. Perhaps nobody can provide evidence that the queue is actually the main delay. Another branch may become more compelling because several people can point to the same recurring approval bottleneck.
You are beginning to learn which explanations deserve more attention.
But you are still not done.
I Often Run Another Iteration
This is one of the most useful parts of 5 Whys for me.
I run it again.
Sometimes I return to the original problem and ask the questions from another angle. Sometimes I take a branch that we did not explore deeply enough. Sometimes I revisit the issue after speaking with people closer to the work. A new observation can make yesterday’s explanation look much weaker.
I may repeat this several times until an aha begins to emerge.
The aha is not necessarily a new fifth answer.
Often it appears because several branches start connecting.
Imagine that one branch tells us employees wait for approval. Another shows that managers frequently reverse independent decisions. A third reveals vague rules about what employees may decide. A fourth shows that escalation has become faster and safer than making a judgment.
Taken separately, these look like different causes.
Taken together, they may suggest a larger explanation:
We say we want frontline ownership, but our working system repeatedly teaches people that independent judgment is risky.
That is an aha.
It explains more of what we are seeing than “employees need more initiative” or “the policy is unclear.”
But the aha is still not a verdict.
It is a stronger hypothesis.
An Aha Is Worth Testing, Not Worshipping
A good explanation earns the right to be tested.
Suppose the team believes low decision authority is a major contributor to slow customer resolution. Leaders raise the limit for common cases, clarify the boundaries, and allow frontline employees to resolve routine problems without waiting for a manager.
Then they watch what happens.
Do resolution times fall?
Do unnecessary escalations decline?
Do employees make sound decisions inside the new boundary?
Do managers continue taking those decisions back?
If performance improves substantially, the explanation gains credibility.
But suppose escalations fall while customer waiting time barely changes. The team looks again and discovers that employees still spend too long searching for account information.
That does not mean the first intervention failed.
It means the first explanation was incomplete.
You removed one source of friction and exposed another.
That is learning.
This is why I prefer to think of 5 Whys as part of a larger loop:
Observe → investigate → form an explanation → change something → follow the proof → learn → ask again.
The purpose of diagnosis is not to reach a point where we can proudly say, “We found the root cause.”
It is to make a better next move.
Ask About the Work, Not Just the Person
There is another reason to use the tool carefully.
“Why?” can quickly sound like accusation.
“Why didn’t you finish?”
“Why didn’t you follow up?”
“Why didn’t you escalate?”
By the fourth or fifth question, the exercise can feel less like inquiry and more like prosecution.
When I am investigating workplace problems, I often vary the question:
“What happened when you tried?”
“What were you waiting for?”
“What made that response seem reasonable at the time?”
“What normally happens when this situation occurs?”
“What would have needed to be different for you to move?”
These questions still help us go deeper, but they keep the inquiry close to the work.
That matters because labels such as “low motivation,” “poor accountability,” and “resistance to change” can sound like causes when they may simply describe behavior we have not yet understood. This is the same reason I argue in Motivation at Work: Before You Blame the People, Diagnose the Work that we should separate what we observed from the explanation we placed on it.
If someone says, “People lack ownership,” ask what people are actually doing. If decisions keep returning to the manager, ask what happens when employees decide independently. The deeper inquiry may reveal something very different from the original label.
Five Is a Prompt, Not a Law
I still call the technique 5 Whys because the name is familiar and the discipline behind it is useful.
But there is nothing magical about the number five.
Sometimes three questions are enough to reveal something worth investigating. Sometimes the first why produces four branches and each branch needs several more questions. Sometimes you reach the fifth why and the answer is still vague:
“Because communication is poor.”
That is not a reason to stop.
What does poor communication mean here?
Who did not know what?
When did the missing information matter?
What normally happens at that point in the work?
You may even finish one iteration, come back tomorrow, and see a question you missed completely.
The discipline is not counting to five.
The discipline is refusing to stop at the first convenient explanation.
Use 5 Whys to Think Better
When I use 5 Whys now, I do not expect a neat chain to hand me the answer.
I start with one observable problem and let plausible causes emerge. When a why has several reasonable answers, I allow the branches to remain visible long enough to learn from them. I compare where the paths converge, where they split, and what assumptions each one contains.
Then I run another iteration when necessary.
I look for the aha that begins connecting several observations into a more useful explanation. And before I treat that explanation as true, I return to the work and look for evidence.
Sometimes the change works.
Sometimes it produces only partial movement.
Sometimes the work tells us that our explanation was wrong.
All three outcomes teach us something.
That is why I no longer think the main value of 5 Whys is reaching the fifth question.
Its value is making us less satisfied with the first answer.
So the next time your group reaches the bottom of a neat five-step chain and someone says, “We found the root cause,” do not rush to close the exercise.
Look at the branches you left behind.
Run another iteration.
Ask what the evidence supports.
And when the aha comes, test it.
The fifth why is not the truth. It is another invitation to learn.
