Originally published on mistakeology.com.

By Dr Tom Hitchcock, founder of Mr Purple (LinkedIn)

The questions after mistakes matter almost as much as the mistake itself. Ask a predictable question and you will often receive a polished answer. Ask a thoughtful one and you may uncover the useful data hiding beneath the wobble. The aim is not to catch somebody out. It is to understand what made sense at the time, what changed, and what the team can now see more clearly.

Why the usual questions after mistakes fall flat

Many learning conversations begin with a perfectly reasonable sounding question: what went wrong? The trouble is that it points everybody towards the wreckage. People start arranging events into a tidy story, often with one obvious cause and one obvious person standing nearest to it.

Reality is usually messier. A decision may have looked sensible with the information available. A workaround may have become normal because the proper route was slow. A handover may have contained the words but not the meaning. The mistake is visible. The conditions around it are quieter.

Try replacing the hunt for a culprit with curiosity about the system. A useful conversation should help people say what they noticed, expected, assumed and adapted to. That is where the learning lives.

Start with what made sense at the time

Hindsight is a confident little creature. Once the outcome is known, earlier warning signs can look enormous. Before the outcome, they may have been faint, familiar or mixed with several other demands.

Begin with questions that return the group to the moment before the wobble:

• What were you trying to achieve?
• What information did you have then?
• What did you expect would happen next?
• What else was competing for your attention?
• Which option looked least risky at the time?

These questions do not excuse poor choices. They reveal the logic behind them. That difference matters. If leaders understand the local logic, they can improve the conditions that shaped the choice. If they only label the choice as careless, the next person may meet the same conditions with a fresher warning poster on the wall.

Use curveball questions to test the tidy story

Every group creates a first explanation. Sometimes it is right. Sometimes it is merely convenient. A curveball question gently nudges the group beyond the first answer without turning the room into a courtroom.

Useful curveballs include:

• What would have needed to be true for this decision to work?
• Where else does this same condition appear without causing trouble?
• Who had a different view, and what were they seeing?
• What nearly prevented the mistake?
• If the outcome had been fine, what would we never have examined?
• Which part of our explanation feels suspiciously neat?

The question about what nearly worked is especially valuable. Near successes reveal safeguards, judgement and good fortune that a simple failure story misses. They show what the team might strengthen, not merely what it should stop.

Separate facts, interpretations and assumptions

A learning conversation can become tangled because facts and interpretations are spoken with equal confidence. Slow the discussion down and sort the material into three simple buckets.

1. Facts: What can we directly observe or verify?
2. Interpretations: What meaning have we given those facts?
3. Assumptions: What did we believe without checking?

For example, a message was sent at a certain time. That is a fact. Somebody ignored it. That is an interpretation until the reason is known. The recipient understood the urgency. That may be an assumption.

This sorting exercise is not about becoming painfully pedantic. It stops a guess from hardening into a verdict. It also makes the next question easier: what would help us check this properly?

Make the room safe enough for an untidy answer

People read the room before they answer the question. If the leader sounds disappointed, rushed or determined to prove a theory, the team will edit itself. The answer becomes safer for the speaker and less useful for the organisation.

Leaders can improve the quality of the data by changing their own behaviour:

• State that the purpose is learning, then behave accordingly.
• Ask one question and allow silence.
• Thank people for awkward information.
• Do not interrupt an answer that challenges your theory.
• Record uncertainties as uncertainties.
• End with a small test, not a grand promise.

This is the practical heart of blameless reporting. Blameless does not mean consequence free or careless. It means the inquiry is designed to discover, rather than to perform outrage. Where a matter involves safety, employment, legal duties or compliance, bring in the appropriately qualified adviser. A learning conversation is useful, but it is not a substitute for specialist advice.

Turn the answers into a small experiment

A good conversation should alter something. Not everything. Just something specific enough to test.

Finish by asking:

• What is one condition we can change?
• What would we expect to notice if the change helps?
• Who will test it?
• When will we look at the result?
• What new risk might the change create?

A small experiment protects the team from the theatre of the enormous action plan. It also produces fresh data. Perhaps a handover prompt improves clarity. Perhaps it adds noise. Either result teaches you something, provided somebody returns to look.

The best questions do not make mistakes disappear. They make mistakes speak. Next time a wobble appears, resist the neat answer and ask what made sense, what was assumed and what nearly worked. Then try the Wobble Map to turn the conversation into a practical picture of what to explore next.

Common questions

What should I ask first after a mistake?

Ask what the person or team was trying to achieve and what information they had at the time. This opens the context before hindsight turns it into a tidy story.

Does a blameless conversation remove accountability?

No. It separates understanding from accusation, so accountability can be informed by what actually happened rather than by the quickest explanation.

How do I challenge an answer without sounding accusatory?

Use curious language and test the shared explanation, not the character of the person. Ask what else might be true or which part of the story feels unusually neat.

What if the mistake involves safety or compliance?

Use the learning questions to understand the conditions, but seek advice from the appropriately qualified safety, legal or compliance professional for any formal obligations.

How many actions should come from a mistake review?

Start with one small, observable experiment where possible. A focused test is easier to assess than a large action plan that nobody revisits.