Root cause analysis
Fixing symptoms keeps a team busy for ever. Root cause analysis digs until the cause that keeps producing the problem is found and proved.
About 5 minutes to read.
Key idea
When something goes wrong, the first explanation that comes to mind is usually a symptom. The parcel was late because the van left late. Why did the van leave late? Because loading finished late. Why? Because the last orders were picked after the cut-off. Asking why, patiently, five times or so, walks the team from the symptom towards a cause they can actually change.
A cause-and-effect diagram, often called a fishbone, keeps the brainstorm broad. The problem sits at the head and the bones are the usual families of causes: people, methods, machines, materials, measurement and environment. The team fills the bones with every plausible cause, then picks the likeliest few to test.
Testing is the step people skip. A cause is a suspect until the data convicts it. Compare the late orders with the on-time ones and see whether the suspected cause really separates them.
Example: wrong deliveries
The blamed new starters
A wholesaler sees a rise in wrong items delivered, and the floor manager blames new starters. The team fishbones the problem and lists eleven possible causes. When they look at a month of errors, new starters account for no more errors than anyone else.
What does separate the errors is the product: two lookalike items with consecutive codes sit side by side on the same shelf. Moving one of them and adding a photo to the pick screen halves the errors. The new starters were innocent all along.