A problem-solving tool becomes dangerous when its simplicity creates more confidence than the evidence deserves.
The Five Whys is one of the most appealing tools in management.
A problem occurs. Ask why.
Take the answer. Ask why again.
Continue until you reach something that looks sufficiently fundamental to be called the root cause.
It is simple, memorable and often extremely useful.
But there is a problem.
Complex organisations are not machines with a single broken component. Their recurring problems emerge from interacting people, incentives, assumptions, processes, information, authority, history and commercial pressures.
In that environment, asking "why?" five times can take you somewhere very deep.
Unfortunately, it can also take you somewhere completely wrong.
Every answer becomes the foundation for the next
Imagine customer complaints have increased.
Why?
Because orders are being delivered late.
Why?
Because Operations is taking too long to process them.
Why?
Because the team is understaffed.
Why?
Because management has not approved additional positions.
Why?
Because cost reduction has been prioritised over service.
There it is.
Five Whys. Root cause identified.
Except perhaps Operations isn't understaffed at all.
Perhaps Sales is routinely accepting customer commitments that cannot be delivered within existing capacity.
Perhaps the forecasting information reaching Operations is unreliable.
Perhaps certain customers are being prioritised informally.
Perhaps an ageing system is creating rework.
Perhaps decision rights around exceptions are unclear.
Perhaps all of these are happening simultaneously.
The problem with the chain is simple: every subsequent why inherits the assumptions embedded in the previous answer.
If the second answer is wrong, questions three, four and five can become increasingly sophisticated explanations of something that was never true.
Depth is not the same as accuracy
This creates a peculiar form of organisational confidence.
The team feels it has investigated deeply because it has travelled several layers beneath the symptom.
But distance from the symptom is not evidence of proximity to the cause.
You can dig five metres down in the wrong place.
You still dug deeply.
This matters enormously with recurring problems because organisations frequently begin the investigation with an existing story about what is happening.
"It's a capability problem."
"The process isn't being followed."
"The system is inadequate."
"Managers aren't holding people accountable."
"Sales keeps overpromising."
Once one of those explanations enters the Five Whys chain, subsequent questions can reinforce it.
The tool begins validating the story rather than investigating the problem.
Complex problems rarely have one neat chain
Five Whys emerged from a manufacturing tradition where causal relationships could often be relatively direct.
A machine stopped.
A component failed.
The component failed because lubrication was inadequate.
Lubrication was inadequate because a pump was malfunctioning.
That kind of causal investigation can be extremely valuable.
But imagine applying the same linear logic to declining customer retention, repeated project overruns, persistent margin erosion, cultural problems, cybersecurity incidents or recurring operational failures spanning several functions.
There may not be one chain.
There may be twenty.
One decision may have several contributing behaviours. Those behaviours may arise from different assumptions. Those assumptions may be reinforced by incentives, leadership reactions, information asymmetries, performance measures and historical experience.
Worse, those factors can reinforce one another.
You are no longer examining a chain.
You are examining a system.
Recurrence is telling you the previous explanation was incomplete
This is where CEOs should become particularly suspicious.
If the organisation has already "solved" the problem several times and it keeps returning, another rapid root-cause exercise may simply produce another plausible explanation.
Then comes another solution.
Another programme.
Another technology investment.
Another restructure.
Another training initiative.
Another accountability mechanism.
The organisation spends more money while confidence in its ability to solve the problem declines.
That is Karamawari: considerable effort without corresponding forward movement.
The recurring problem is providing valuable information.
It is telling you that something capable of recreating the problem survived the previous solutions.
That deserves a broader investigation than simply asking why again.
Ask "what else?" as often as "why?"
One useful adjustment is surprisingly simple.
After asking why something happened, ask:
What else could explain this?
Then look for evidence capable of distinguishing between those explanations.
If Operations says delays result from insufficient resources, don't immediately make understaffing the next link in the chain.
Test it.
When do delays occur?
When don't they occur?
Which teams experience them?
Which customers?
What changes under pressure?
What information was available when decisions were made?
What incentives influenced those decisions?
What assumptions were people making?
What conditions made those assumptions reasonable?
Suddenly the investigation branches rather than tunnels.
That may feel messier.
Good.
The organisation itself is messy.
The tool isn't the problem
The lesson is not to abandon Five Whys.
The mistake is asking a simple tool to carry more diagnostic weight than it was designed to bear.
Use it when causal relationships are sufficiently bounded and evidence can validate each step.
Become cautious when problems cross functions, involve human judgement, change under pressure, have several plausible explanations or have survived previous attempts to eliminate them.
And recurrence itself should be one of the strongest warning signs.
If your organisation has asked "why?" repeatedly, implemented reasonable answers and watched the same expensive problem return, the next question may not be another why.
How would you know whether you are getting closer to the real cause, or simply getting better at explaining the wrong one?
0 Comments