Contents
Use this when a group is stuck in circular debate, when the loudest voice is drowning out the most important perspective, or when a decision feels simultaneously obvious and uncomfortable. Six Thinking Hats separates thinking into six distinct modes — facts, feelings, risks, benefits, creativity, and process — so that everyone examines the same angle at the same time, replacing adversarial argument with parallel exploration.
Section 1
What This Tool Does
Watch any leadership team debate a consequential decision. Within ninety seconds, the pattern reveals itself. The CFO raises a risk. The head of product counters with an opportunity. The CEO shares a gut feeling dressed up as strategy. Someone cites a data point that supports their pre-existing position. Someone else cites a different data point that supports theirs. The conversation isn't a conversation — it's six simultaneous monologues, each participant locked into a cognitive mode they've occupied since before the meeting started. The optimist argues for the opportunity. The sceptic argues against it. The analyst wants more data. The creative wants to brainstorm alternatives. Nobody is wrong. But nobody is thinking together, either.
Edward de Bono diagnosed this problem in the early 1980s while consulting with corporations across Europe and observing a pattern so consistent it amounted to a law: in unstructured group discussion, people don't think — they defend. Each person adopts a position early, then spends the rest of the meeting marshalling arguments for that position and attacking arguments against it. This is adversarial thinking, borrowed from Western legal tradition. It's excellent for courtrooms. It's terrible for decisions where the goal is to see the full landscape before committing.
De Bono's intervention was structural, not motivational. He didn't ask people to be more open-minded. He changed the rules of the conversation. Six Thinking Hats assigns a coloured hat to each mode of thinking: White for facts and data, Red for emotions and intuition, Black for risks and caution, Yellow for benefits and optimism, Green for creative alternatives, Blue for process management. The critical move: everyone wears the same hat at the same time. When the group is wearing the Black Hat, everyone — including the most enthusiastic advocate — looks for risks. When the group switches to Yellow, everyone — including the most committed sceptic — looks for benefits. The adversarial dynamic dissolves because there are no positions to defend. There is only a shared direction of attention.
The cognitive shift this creates is more profound than it appears. Most people believe they're capable of balanced thinking — weighing pros and cons, considering emotions alongside data. They're not. Kahneman's research on cognitive load demonstrates that the brain struggles to hold multiple evaluative frames simultaneously. You can analyse data or process an emotional reaction, but doing both at once degrades both. De Bono's hats don't ask you to think about everything at once. They ask you to think about one thing at a time, thoroughly, before moving to the next. Sequential depth instead of simultaneous shallowness. That's the whole trick.
The result, when the tool is used properly, is a decision that has been examined from six angles by the entire group — not six angles distributed across six individuals who each examined one. The difference matters enormously. A risk identified by the person who also identified the opportunity carries different weight than a risk raised by the designated sceptic. When the optimist finds the flaw, the room listens.
Section 2
How to Use It — Step by Step
Instructions on the left. Worked example — "Should our B2B SaaS company (ARR $18M, 140 employees) acquire a smaller competitor ($2M ARR, 22 employees) to consolidate market share in the mid-market segment?" — on the right.
Step 1 — Frame
Set the question and sequence the hats (Blue Hat)
The Blue Hat is the meta-hat — it controls the process. Start every session here. Define the decision question in one sentence. Then choose the hat sequence. There's no single correct order, but a reliable default is: Blue → White → Red → Yellow → Black → Green → Blue. Open with facts, surface gut reactions early (before they go underground and contaminate the analytical hats), explore upside before downside (so the Black Hat critiques something specific rather than killing a vague idea), then generate alternatives. Close with Blue to synthesise. Allocate time per hat — typically 5–10 minutes each for a 60-minute session. Appoint a facilitator who enforces hat discipline: if someone raises a risk during Yellow Hat time, they're redirected, not silenced. "Hold that — we'll get to Black Hat in eight minutes."
Worked example
Acquisition decision — framing
The CEO opens: "We're evaluating whether to acquire CompactCRM. Blue Hat first — here's our sequence: White, Red, Yellow, Black, Green, then Blue to close. Ten minutes per hat. I'll facilitate. The question: should we acquire CompactCRM for the proposed $6.5M, and if not as structured, under what conditions might we proceed?" The facilitator writes the question on the whiteboard. Everyone can see it. No one is allowed to answer it yet.
Step 2 — Gather
Lay out the facts without interpretation (White Hat)
White Hat thinking is strictly informational. What do we know? What don't we know? What data would we need to verify our assumptions? No opinions, no judgments, no "and that means..." — just facts and identified gaps. This is harder than it sounds. Most people can't state a fact without immediately interpreting it. The facilitator's job is to catch interpretations and redirect: "That's a Black Hat thought — park it. What's the underlying data point?" Document everything visibly. Two columns work well: "Known" and "Unknown / Need to verify."
Worked example
Acquisition decision — facts
Known: CompactCRM has $2M ARR, 85% gross retention, 22 employees (14 engineering, 4 sales, 4 ops). Their product covers 60% of our feature set with a different UI architecture. They have 340 customers, 40 of which overlap with our prospect pipeline. The asking price is $6.5M (3.25× ARR). Our cash position is $11M with $3M quarterly burn. Unknown: Their net revenue retention (expansion vs. contraction). Technical debt level. Key-person risk — does the CTO stay? IP ownership clarity. Customer contract transferability. Integration timeline and cost. Each unknown gets flagged for due diligence.
Step 3 — React
Surface emotions and intuitions without justification (Red Hat)
Red Hat is the most counterintuitive step for analytical teams — and the most important one to protect. Each person states how they feel about the decision. No reasons. No data. No "I feel concerned because..." Just: "I feel excited." "I feel uneasy." "Something about this feels rushed." The prohibition on justification is essential. The moment someone explains their feeling, the group starts debating the explanation instead of registering the signal. Feelings are data — they encode pattern recognition that hasn't yet been articulated. A room where four of seven people feel uneasy is a room that has detected something the White Hat data hasn't surfaced yet. Record the feelings. Don't discuss them. Move on.
Worked example
Acquisition decision — emotions
VP Engineering: "Nervous. Integrations always take longer than we think." Head of Sales: "Excited — those 40 overlapping prospects could close faster." CFO: "Cautious. The price feels high for the retention numbers." CPO: "Curious but sceptical." CEO: "Energised but aware I might be anchored on the narrative." CTO: "Dread. Honest dread." Two people excited, four some shade of cautious or negative. That asymmetry is noted. No one debates it.
Step 4 — Evaluate
Explore benefits, then risks, then alternatives (Yellow → Black → Green Hats)
Yellow Hat: What's the best realistic outcome? What value does this create that we can't create organically? What strategic advantages does it unlock? Push the optimists and the sceptics alike to find genuine upside. Black Hat: What could go wrong? What are the risks, the costs, the second-order consequences? Where are the assumptions most fragile? This is not pessimism — it's critical examination. The goal is to identify every failure mode, not to kill the idea. Green Hat: What alternatives exist? What modifications to the deal would address the Black Hat concerns while preserving the Yellow Hat benefits? What if we did something entirely different? Green Hat is where creative options emerge — partial acquisitions, licensing deals, acqui-hires, partnerships. Protect this space from premature evaluation.
Worked example
Acquisition decision — evaluate
Yellow: Acquiring 340 customers accelerates mid-market penetration by 12–18 months. Engineering team fills our mobile gap. Removes a competitor from the market. 40 overlapping prospects become warm leads. Black: Integration could consume engineering for two quarters, stalling our roadmap. 85% gross retention means 15% annual churn — are we buying a leaky bucket? $6.5M is 59% of cash reserves; one bad quarter post-acquisition and we're fundraising from a position of weakness. CTO departure risk is high (Red Hat "dread" may be tracking this). Green: Counter-offer at $4.5M with $2M earnout tied to retention. Acqui-hire the engineering team only and let the product sunset. License their mobile technology instead of acquiring. Propose a 12-month partnership to test integration before committing.
Step 5 — Synthesise
Consolidate findings and define the decision path (Blue Hat)
Return to Blue Hat. The facilitator summarises what each hat revealed: the facts and gaps (White), the emotional signals (Red), the upside case (Yellow), the risk landscape (Black), and the alternative options (Green). The group doesn't need to decide in this session — often the best outcome is a clear next step rather than a premature commitment. What due diligence is required to close the White Hat gaps? Which Black Hat risks are dealbreakers versus manageable? Which Green Hat alternative deserves further exploration? The Blue Hat synthesis should produce a one-page summary: the question, key findings per hat, and the agreed next action.
Worked example
Acquisition decision — synthesis
Blue Hat summary: "The strategic case (Yellow) is real but not urgent — we could build the mobile capability in 9 months. The financial risk (Black) is significant given our cash position. The emotional signal (Red) skews cautious, with the CTO's reaction flagging integration risk that the data hasn't quantified yet. Recommended path: pursue the acqui-hire variant (Green) at a lower price point. Commission a 2-week technical due diligence sprint to assess integration cost and key-person risk before any term sheet. Reconvene in 3 weeks with White Hat gaps closed." The team leaves with clarity, not consensus on a premature answer.
Section 3
When It Works Best
✓
Ideal Conditions for Six Thinking Hats
| Dimension | Best fit |
|---|---|
| Group dynamics | Teams where one or two dominant voices consistently steer the conversation. The hat structure equalises participation by making every person engage with every mode — the sceptic must find benefits, the optimist must find risks. Most effective in groups of 4–8 where interpersonal dynamics are strong enough to distort unstructured discussion. |
| Decision complexity | Multi-dimensional decisions where the analytical, emotional, creative, and risk dimensions all matter but tend to get collapsed into a single debate. M&A, market entry, major product pivots, organisational restructuring — decisions where the cost of missing a perspective exceeds the cost of a 60-minute structured session. |
| Emotional charge | Decisions where people have strong feelings they're reluctant to express in analytical settings. The Red Hat legitimises emotional input by giving it an explicit, time-boxed slot. Teams that suppress emotions don't eliminate them — they just push them underground where they distort the "rational" analysis without anyone acknowledging it. |
| Cognitive diversity | Cross-functional teams where each member naturally gravitates to a different thinking mode. The hats force the engineer to think creatively and the designer to think about risk — stretching each person beyond their default mode. Homogeneous teams benefit less because the natural diversity of perspectives is lower. |
| Time pressure | Paradoxically, the tool works well under moderate time pressure. The hat structure prevents the meeting from spiralling into unstructured debate. A 60-minute Six Hats session often produces more actionable output than a 3-hour free-form discussion because the structure eliminates repetition, position-defending, and tangential arguments. |
| Recurring decisions | Teams that make similar decisions repeatedly (investment committees, product prioritisation, hiring panels) benefit from adopting the hats as a standing protocol. The shared vocabulary — "let's do a quick Red Hat check" — becomes a cultural shorthand that persists beyond formal sessions. |
Section 4
When It Breaks Down
⚠
Failure Modes
| Failure pattern | What goes wrong | What to use instead |
|---|---|---|
| Hat contamination | Participants smuggle their preferred thinking mode into the wrong hat. The sceptic frames risks as "facts" during White Hat. The advocate frames benefits as "feelings" during Red Hat. Without a disciplined facilitator who catches and redirects these violations in real time, the hats become labels on the same unstructured argument. | Assign a dedicated facilitator whose only job is hat discipline — they don't contribute content, only enforce the mode. |
| Performative participation | People go through the motions without genuine cognitive switching. They say something during Yellow Hat time because they're supposed to, but their actual assessment hasn't shifted. This is especially common in hierarchical cultures where disagreeing with the boss — even under the protection of a hat — feels unsafe. | Combine with anonymous written input (silent brainstorming per hat) before group discussion to surface genuine perspectives. |
| Insufficient data for White Hat | If the team lacks basic facts about the decision, the White Hat phase exposes the gap but can't fill it. The remaining hats then operate on assumptions rather than information. The session feels productive but the foundation is sand. | Run a pre-session data gathering sprint. Use Issue Trees or 5 Whys to identify what information is actually needed before convening the hats session. |
| Trivial decisions | Applying six modes of structured thinking to a decision that doesn't warrant it. Choosing a project management tool doesn't need a Green Hat creative session. The overhead of the process exceeds the value of the decision, and the team learns to associate the hats with bureaucratic waste rather than genuine insight. | Reserve for decisions above a clear threshold — irreversible commitments, decisions involving >10% of budget, or choices where the team is genuinely split. |
| Black Hat dominance | Risk-averse cultures spend disproportionate time and energy on the Black Hat, producing exhaustive risk catalogues that paralyse action. The Yellow and Green Hats get compressed or skipped. The tool becomes a sophisticated mechanism for saying no. | Enforce strict time parity across hats. If Black Hat gets 10 minutes, Yellow Hat gets 10 minutes. No exceptions. |
| Solo use without adaptation | The tool was designed for groups. Using it alone requires genuine cognitive discipline — most individuals can't fully suppress their dominant thinking mode even when deliberately trying to switch hats. The solo version tends to produce a more articulate version of whatever the person already believed. | For solo decisions, pair with Inversion or Pre-Mortem to force adversarial self-examination. Or write out each hat's perspective in full sentences — the act of writing forces more rigorous mode-switching than thinking alone. |
The most dangerous failure mode is hat contamination, because it's invisible. A session can feel structured, disciplined, even productive — and still produce a biased outcome if participants are encoding their existing positions into whichever hat is currently active. The CFO who frames "I think the price is too high" as a White Hat observation ("the multiple is 3.25× ARR") has smuggled a Black Hat judgment into the data phase. The rest of the group now processes that number as a fact rather than an interpretation, and the anchoring effect shapes every subsequent hat.
The protection is a facilitator who understands the difference between data and interpretation, between feeling and analysis, between risk identification and pessimism. This person needs to be willing to interrupt senior leaders mid-sentence. "That's an interpretation — park it for Black Hat. What's the raw number?" If your culture can't support that kind of facilitation, the hats will be decorative.
Section 5
Visual Explanation
Six Thinking Hats — applied to the B2B SaaS acquisition decision. Each hat represents a distinct cognitive mode; the group cycles through all six sequentially.
Section 6
Pairs With
The Six Thinking Hats structures how a group thinks. It doesn't structure what they think about, what data they bring, or what they do with the output. The tools below fill those gaps.
Use before
First Principles Thinking
Before convening the hats, decompose the decision into its fundamental assumptions. This gives the White Hat phase real material to work with — not just surface-level data, but the underlying premises the decision rests on. A hats session built on unexamined assumptions produces six perspectives on the wrong foundation.
Use before
Reframing
The hats examine a decision from six angles, but all six angles are pointed at the same question. If the question is wrong, the analysis is wasted. Reframing challenges the decision frame itself — "Should we acquire them?" might become "How do we close our mobile gap?" — before the hats session begins.
Use after
Pre-Mortem
The Black Hat identifies risks. The Pre-Mortem goes further: it assumes the decision has already failed and asks why. Run a Pre-Mortem on the option that emerges from the hats session to stress-test the risks the Black Hat surfaced — and to find the ones it missed.
Use after
Decision Matrix
When the Green Hat generates multiple alternatives, the Decision Matrix provides a structured way to compare them against weighted criteria. The hats surface options and perspectives; the matrix forces a quantified comparison.
Mental model
Ladder of Inference
The Ladder of Inference explains how people leap from data to conclusions through layers of selection, interpretation, and assumption. Understanding this model helps facilitators catch hat contamination in real time — recognising when a participant has climbed the ladder and is presenting a conclusion as if it were data.
Mental model
Second-Order Thinking
The hats tend to surface first-order effects: direct benefits (Yellow), direct risks (Black). Second-Order Thinking pushes each hat's output one step further — "and then what?" The acquisition removes a competitor (Yellow, first-order), which signals to the market that consolidation is underway, attracting a larger competitor to enter (second-order).
Section 7
Real-World Application
Prudential Insurance — restructuring product development decisions
The scenario
In the early 1990s, Prudential Insurance adopted Six Thinking Hats across its product development and strategic planning functions after Edward de Bono conducted training sessions with senior leadership. The immediate catalyst was a pattern the executive team recognised in their own behaviour: product launch decisions were being dominated by two camps — actuaries who focused almost exclusively on risk (Black Hat thinking, though they didn't call it that) and sales leaders who focused almost exclusively on market opportunity (Yellow Hat thinking). Meetings devolved into positional warfare. Products either launched late because risk concerns kept sending them back for revision, or launched prematurely because commercial pressure overrode actuarial caution. The middle ground — where a product is appropriately priced for its risk profile and timed for its market — was rarely found through debate.
How the tool appliedPrudential implemented the hats as a standing protocol for product review meetings. The critical structural change: actuaries were required to spend Yellow Hat time articulating the commercial upside of a product, and sales leaders were required to spend Black Hat time articulating the risk exposure. This wasn't symbolic. Each participant had to produce substantive contributions in their non-native mode. The White Hat phase was formalised with a pre-meeting data pack — market sizing, claims history, competitive pricing, regulatory requirements — so the session didn't waste time on information that should have been distributed in advance. Red Hat was initially the most resisted phase; insurance executives were uncomfortable stating feelings without justification. Within three months, it became the most valued — gut reactions from experienced underwriters frequently flagged risks that the quantitative models hadn't captured.
What it surfacedThe Green Hat phase proved unexpectedly productive. In the old adversarial format, creative alternatives rarely emerged because the meeting structure was binary: approve or reject. The Green Hat created a dedicated space for "what if we structured it differently?" One product — a long-term care insurance offering — had been rejected twice in traditional review meetings because the risk profile was unacceptable at the proposed price point. During a Green Hat session, a junior actuary suggested a tiered benefit structure that reduced the insurer's tail risk while preserving the product's market appeal. The product launched with the modified structure and became one of Prudential's stronger performers in the category.
The non-obvious factorWhat made Prudential's adoption work where many corporate implementations fail was the requirement for substantive cross-mode contribution. Most organisations adopt the hats as a discussion format — everyone talks during each hat phase, but the natural sceptics still dominate Black Hat and the natural optimists still dominate Yellow. Prudential explicitly required people to contribute outside their default mode, and evaluated the quality of those contributions. This forced genuine cognitive switching rather than performative compliance. The actuaries didn't just tolerate Yellow Hat time; they had to be good at it. That requirement — uncomfortable, initially resisted — was the mechanism that broke the adversarial pattern.
Section 8
Analyst's Take
Faster Than Normal — Editorial View
Six Thinking Hats has a branding problem. The coloured hats, the slightly playful metaphor, the association with corporate training workshops — all of it makes the tool easy to dismiss as lightweight. Serious operators reach for decision matrices and scenario models. The hats feel like something from an offsite icebreaker. This is a mistake. The tool solves a problem that no amount of analytical sophistication can address: the tendency of groups to argue rather than think. A decision matrix is useless if the team populating it is locked in positional combat. A scenario model is useless if the assumptions feeding it were never examined from more than one cognitive angle. The hats don't replace analytical tools — they create the conditions under which analytical tools actually work.
The failure mode I see most often is what I'd call "hat tourism." The team cycles through all six hats in thirty minutes, spending five minutes per hat, generating a few bullet points in each mode, and declaring the exercise complete. The output looks balanced. It isn't. Five minutes of Yellow Hat thinking doesn't surface the non-obvious benefits — it surfaces the obvious ones that everyone already knew. Five minutes of Green Hat doesn't produce creative alternatives — it produces the first idea that comes to mind, which is almost never the best one. The hats work through depth, not breadth. Better to spend a full session on just three hats — White, Black, and Green, say — and go deep, than to race through all six and go shallow. The six-hat sequence is a menu, not a mandate.
The highest-leverage modification: separate the Red Hat from the rest and run it first, before any data is shared. Ask each person to write down their gut reaction to the decision question before the White Hat data pack is distributed. Seal the responses. Then run the full session. At the end, open the sealed Red Hat responses and compare them to the session's conclusion. When the data-informed conclusion diverges from the initial gut reactions, you've learned something important — either the data changed minds (good) or the data was marshalled to rationalise what people already felt (dangerous). This before-and-after comparison turns the Red Hat from a single data point into a calibration tool for the entire session's intellectual honesty.
Section 9
Top Resources
01
Six Thinking Hats — Edward de Bono (1985)
Primary sourceThe original and still the essential reference. De Bono's writing is deliberately accessible — some readers mistake this for lack of depth. It isn't. The book's core chapters on each hat contain nuances that most summaries miss entirely, particularly his guidance on when to use abbreviated sequences versus full six-hat rotations, and his distinction between "initial" and "evolved" applications of the Red Hat. Read the revised 1999 edition, which incorporates fifteen years of corporate implementation feedback.
02
Book
The scientific foundation for why the hats work. Kahneman's research on System 1 and System 2 thinking, anchoring effects, and the affect heuristic explains precisely why unstructured group discussion produces biased decisions — and why forcing sequential, single-mode thinking is a genuine cognitive intervention. Read Part I (Heuristics and Biases) and Part III (Overconfidence) for the most relevant material.
03
Lateral Thinking: Creativity Step by Step — Edward de Bono (1970)
BookThe intellectual precursor to Six Thinking Hats. De Bono's earlier work on lateral thinking — the deliberate generation of alternative perspectives — provides the theoretical framework that the hats operationalise. The Green Hat is essentially lateral thinking given a structured slot in a group process. Understanding the underlying theory makes facilitation significantly more effective, particularly when teams struggle with the creative mode.
04
Book
Lafley and Martin's strategy framework at Procter & Gamble embeds a version of parallel thinking into strategic choice-making — their "reverse engineering" of strategic options requires teams to build the strongest possible case for each option before evaluating. The structural similarity to Yellow Hat thinking applied to multiple alternatives is instructive. Chapter 7 on "keeping two ideas in your head at once" is the closest modern strategy text gets to de Bono's core insight.
05
Book
Not a book about the hats, but a masterclass in the problem they solve. Horowitz's accounts of wartime CEO decisions — layoffs, pivots, near-death fundraises — illustrate what happens when leaders collapse multiple thinking modes into a single agonised judgment. His chapter on "the struggle" is essentially a case study in what happens when Red Hat signals (fear, dread, loneliness) go unexamined and contaminate every other mode of analysis. Read it as a cautionary tale about the cost of not separating your thinking.
Decision Tools Library — Browse by phase
Why this matters next
mental modelsSecond-Order Thinking
First Principles Thinking applied the Second-Order Thinking mental model
mental modelsFirst Principles Thinking
First Principles Thinking applied the First Principles Thinking mental model
mental modelsLeverage
First Principles Thinking applied the Leverage mental model
mental modelsTechnical Debt
First Principles Thinking applied the Technical Debt mental model
mental models5 Whys
First Principles Thinking applied the 5 Whys mental model
mental modelsNarrative
First Principles Thinking applied the Narrative mental model
Continue exploring
DMCADTSP
Decision tool
Decision Matrix
Score options across weighted factors to identify a clear winner
Decision tool
Cost-Benefit Analysis
Quantify the economic trade-offs of each option to determine net value
Decision tool
Decision Tree
Map sequential outcomes with probabilities and expected values to find the highe
Decision tool
Scenario Planning
Construct multiple plausible futures and test how each option performs across th
More like this, in your inbox
I send a newsletter every week — free, no spam, unsubscribe anytime.