What the Six Hats are — and the problem they solve
Watch a strategy meeting fail and you'll usually see mode collision: one person cites figures, another voices doubt, a third defends the idea, a fourth proposes an alternative — simultaneously. Every contribution is legitimate; the collision wastes them all. Edward de Bono's 1985 book Six Thinking Hats offered a disarmingly literal fix: do one mode at a time, everyone together — parallel thinking instead of adversarial thinking.
The six modes: White Hat — facts, data, and (crucially) named information gaps; Red Hat — feelings and intuition, voiced without justification required; Black Hat — caution, risks, why it might fail; Yellow Hat — benefits, value, why it could work; Green Hat — creativity, alternatives, provocations; Blue Hat — process control: setting the objective, choosing the sequence, synthesizing the output.
The psychological engineering is the underrated part. When the whole room wears Black together, criticism is the assigned task — nobody is attacking anybody's baby. When the room wears Green, no idea gets shot down mid-generation, because critique is simply not what's happening yet. The hats license modes that organizational culture suppresses (junior skepticism, senior uncertainty, anyone's gut feeling) by making them impersonal. De Bono (1933–2021), the Maltese physician-psychologist who coined 'lateral thinking', spent a career arguing thinking is a trainable skill; the hats are his most operational proof. Context in the toolbox: decision-making models and the chooser's guide.
When to use it — and when not to
Six Hats earns its keep when:
- ✓A decision needs full-spectrum examination — strategic choices and complex trade-offs where facts, risks, benefits and alternatives all deserve dedicated attention rather than a brawl.
- ✓The group has stuck dynamics. The chronic critic, the unsinkable optimist, the silent intuitive — hats rotate everyone through every mode, breaking personality lock-in.
- ✓Pre-mortem energy is needed without the gloom. A disciplined Black Hat pass surfaces the objections before the board meeting does — with Yellow's balance already on the record.
And its failure modes:
- ✗Black Hat dominance. Risk-minded groups overstay the comfortable hat. Countermeasure: Yellow before Black for fragile ideas, and time-box every hat.
- ✗Skipping Blue. No framing at the start, no synthesis at the end — six modes of output and no decision. Always bookend.
- ✗Ritual rigidity. The full sequence for a small decision is overhead theater. Quick calls need three hats (White–Yellow–Black) and ten minutes.
- ✗Session amnesia. The subtlest failure: brilliant parallel thinking that evaporates when the meeting ends, because flipchart photos are not a record. That is the argument-tree connection below.
Step by step, with a worked example
Illustrative scenario: an invented 40-person product team deciding whether to sunset its legacy mobile app. A 60-minute facilitated pass:
- 1Blue (5 min): frame. The facilitator states the question — "Do we sunset the legacy app in Q2?" — the sequence, and the time-boxes. Decision, not discussion topic.
- 2White (10 min): facts and gaps. Usage down to 8% of sessions; maintenance consuming one engineer; enterprise contract references 'mobile access' ambiguously. Named gap: no data on which customers those 8% are. Gaps are White-Hat output, not embarrassments.
- 3Green (10 min): alternatives. Beyond kill-or-keep: sunset with a migration wizard; hand the app to a partner; wrap the web app natively; sunset for consumers but keep an enterprise build.
- 4Yellow (10 min): benefits — before Black, deliberately. For the leading options: engineering capacity freed for the roadmap; one codebase; the enterprise-build variant preserves the contract story.
- 5Black (10 min): risks, hard. The 8% might include the two largest accounts (the White-Hat gap now has teeth); app-store reviews during sunset can bleed into brand search; the contract ambiguity is legal exposure until read.
- 6Red (5 min): gut checks, no justification. Three people feel the Q2 date is arbitrary urgency. That feeling is data about the room.
- 7Blue (10 min): synthesize. Decision: resolve the White-Hat gap and the contract read within two weeks; default to the enterprise-build sunset variant unless the gap-data objects. Owner and date named.
The hats as an argument tree
In decision-quality terms, Six Hats feeds alternatives (Green is a dedicated engine for them) and disciplines the reasoning process itself — it is the rare framework aimed squarely at how a group thinks. Its output, though, is only as durable as the flipchart. Mapping hats to an argument tree preserves the structure the session created:
Blue → root claim & synthesis
The framing question becomes the root; the closing synthesis becomes the recorded decision on it, with the reasoning attached.
White → evidence · Yellow/Black → pros & cons
Facts and named gaps attach as evidence; every Yellow contribution is a supporting argument, every Black one an attacking argument — typed exactly as the hats typed them.
Green → new options and compromises
Alternatives enter as sibling options under the root; hybrid proposals (the enterprise-build variant) are compromise nodes with their own pros and cons.
Red → labeled intuition, ratings capture the room
Gut checks enter explicitly labeled as intuition, and the ratings pass captures where the whole room stands — the Red Hat, quantified.
Run this way — start from the Six Thinking Hats meeting template — the session's structure survives it. The Black-Hat objection that killed the naive plan is findable in March; the follow-up on the White-Hat gap lands as new evidence on the exact node that demanded it; and challenges to any claim continue asynchronously through structured chains after the meeting ends.
The hats discipline the thinking; the argument tree keeps it. Six Hats supplies alternatives and process — the tree supplies the sound reasoning record that outlives the room. See decision quality.
Six Hats vs the alternatives
| If your question is… | Reach for | Why not Six Hats |
|---|---|---|
| Pure risk interrogation of one plan | A pre-mortem — see quick strategies and the facilitation tutorial | A full-spectrum pass dilutes the dedicated failure hunt |
| Weighing framed options on their merits | Cost-benefit analysis or a decision tree | Hats structure modes of thought, not valuation |
| Bias patterns are the specific worry | Prospect theory + bias guides | Hats mitigate some biases procedurally but don't diagnose them |
| Deciding who decides, not how to think | RAPID vs DACI vs RACI | Different question entirely — authority, not cognition |