5 Whys

What it means

The 5 Whys is a simple technique for finding the root cause of a problem. It consists of asking why repeatedly until you get from the visible symptom to the real reason. Five is only a rule of thumb; the point is not to stop at the first explanation.

The method comes from the Toyota Production System and is associated with Taiichi Ohno, who used it to investigate production problems. It is now part of lean and kaizen approaches and is widely used in services, administration and IT.

Its strength is simplicity. It needs no special tools or data, only a team that knows the problem and a willingness to dig deeper. It often reveals that a technical error stems from a missing procedure, unclear responsibility or a poorly designed process.

How it works in practice

An example from accounting: why did we pay a penalty? Because the VAT return was filed late. Why? Invoices were missing. Why? The client sent them after the deadline. Why? They did not know the deadline. Why? We never clearly told them. The fix is a regular reminder, not just a reprimand for an employee.

The method complements the Ishikawa diagram well. The diagram maps possible causes across different areas, and the 5 Whys then drills down into the most likely ones. Both are often used in root cause analysis of complaints and in an 8D report.

The outcome should be an action that removes the root cause and a check that the problem has really gone away. In this way the method fits the PDCA cycle: plan the change, do it, check the effect and standardise the change, for example in a procedure or checklist.

What to watch out for

A common mistake is stopping at blaming a particular person, for example with the answer that an employee was careless. That is just another symptom; ask why the system allowed the error and look for ways to prevent it in future regardless of who does the work.

For complex problems with several simultaneous causes, linear questioning can oversimplify reality and lead to only one of them. In that case it helps to follow several branches at once, verify answers with data and not rely solely on the participants' assumptions.