Making the invisible visible

One problem, four ways of looking

The iceberg model helps us look beneath the visible event or occurrence. Here is an illustrative delivery example—not a diagnosis:

• Event—what happened? A customer’s order arrives late.
• Pattern—what happens over time? Late deliveries recur near the end of each month.
• Structure—how is the work organised? Sales promises dates without checking delivery capacity. Orders pile up before the monthly target deadline.
• Mindset or belief—what assumptions support this? “We must promise the date the customer wants, or we’ll lose the sale.”

The belief may make sense to the people holding it, yet the resulting promises may undermine reliable delivery. That is something to investigate, not assume.

Three practical ways to look deeper

• Sketch the pattern over time. Move from “what happened today?” to “what has been happening over time?” Is the problem recurring, developing or staying much the same? A simple Behaviour Over Time sketch helps the team compare impressions and decide what to investigate. Check those impressions against recorded data; the shape alone does not establish a cause. How to make and use a BOT sketch →
• Follow the work with the people doing it. Trace an order from promise to delivery. Where does it wait? What information, approval or capacity is missing? Include dependencies on other teams or organisations and outside influences such as changing demand.
• Ask what keeps the arrangements in place. What do targets, rules and rewards encourage? Which beliefs are treated as facts? In this example: do customers really leave if offered a realistic date? Check the assumption with evidence.

There may be several interacting causes, not one hidden “root cause”. You do not need a detected chart change before investigating a recurring problem—or an urgent safety concern.

What’s your next useful question?

What keeps happening—and what in the way we organise the work might help explain it?

Choose one recurring problem. Bring together the people who do the work and those affected by it. Agree one explanation to check and what evidence would help.

Then return to the three questions: right problem, right intervention, right measures.