The six hats
- White — facts. What we know, what we don’t know, what data we would need. No opinions.
- Red — feelings. What our gut says, what we like and what we don’t, with no need to justify it.
- Black — caution. The risks, the weak points, why it might not work.
- Yellow — optimism. The benefits, the value, why it might work.
- Green — creativity. Alternatives, new ideas, different ways of seeing the problem.
- Blue — process. What we are talking about, in what order, what we have decided.
The point is not to give each person a hat, but to have everyone wear the same hat at the same time. When everyone looks at the risks together, the pessimist is no longer “the one holding things up”; when everyone looks for the benefits, even the skeptic contributes.
When to use it
It is useful for decisions with many facets — a repositioning, a line launch, a choice between two concepts — and when the group tends to polarize. It is less useful for simple problems, where a structured six-part discussion is overkill.
How to do it
1. Open with blue
Define the problem in one sentence and decide the sequence of hats. A common sequence: white, green, yellow, black, red, blue — but it should be adapted to the purpose.
2. One hat at a time, timed
A few minutes per hat. The facilitator brings the discussion back on track when someone switches hats without noticing — a criticism during yellow, for example.
3. Write everything down
Each hat produces a list. At the end you have facts, feelings, risks, benefits and ideas, kept separate: far easier to read than meeting minutes.
4. Close with blue
What we decided, what still needs checking, who does what. The white hat has often surfaced missing data: that becomes the next steps.
The most common mistakes
- One hat per person. Everyone defends their role, and you are back to the usual discussion.
- Skipping red. Feelings are there anyway: better to state them than let them pass as arguments.
- Black too early. If the risks come before the ideas, the ideas never arrive.
- No conclusion. Without the final blue, the method produces lists but not decisions.
