Sign inGet started
Skills / Leadership-strategy

Cynefin Framework

In a complicated context, at least one right answer exists. In a complex context, however, right answers can't be ferreted out. Instead, instructive patterns emerge if the leader conducts experiments that can safely fail.

The Cynefin Framework (pronounced 'kuh-NEV-in', Welsh for 'place of multiple belongings') is a sense-making tool developed by Dave Snowden that helps leaders classify problems into five distinct domains — Clear, Complicated, Complex, Chaotic, and Disorder — before choosing how to respond. Each domain has its own response pattern, and applying the wrong one is the most common and costly mistake in organizational decision-making. Used by executives, consultants, Agile practitioners, and policy makers worldwide, Cynefin prevents the trap of imposing false order on situations that can only be navigated through emergence.

By Dave Snowden · Free
Specimen 01 · Live diagnosisCynefin Framework
Input

“We're a mid-size software company and we just lost our biggest customer — 40% of revenue. Leadership is split on whether to double down on our…”

Diagnosis
Let's use Cynefin to figure out what kind of situation you're actually in — because the right response depends entirely on that classification, and the most expensive mistakes happen when leaders misread the domain.
Full transcript ↓
Calibrated referenceagent-skills.ai
The gap

Stop treating complex problems like complicated ones — they demand probes, not experts

Cynefin organizes all problems and decisions into five sense-making domains, each defined by a different relationship between cause and effect. In the Clear domain, cause-effect is predictable: Sense, Categorize, Respond using best practices — but complacency risks a sudden 'cliff' collapse into Chaos. In Complicated, cause-effect exists but requires expert analysis: Sense, Analyze, Respond using good practices. In Complex — where Snowden argues most modern organizational challenges actually live — cause-effect can only be perceived in retrospect: Probe, Sense, Respond through safe-to-fail experiments and emergent practices. In Chaotic, no cause-effect holds: Act, Sense, Respond to stabilize the crisis and invent novel practice. Disorder, the central domain, is the danger zone where practitioners don't know which domain they're in and default to cognitive habit. The framework's defining insight is that domain misclassification — especially treating Complex situations as Complicated — is the primary failure mode in organizational decision-making. Associated practitioner tools include SenseMaker (micro-narrative analysis), Ritual Dissent (adversarial stress-testing), and Estuarine Mapping (constraint analysis in complex systems).

The problem

Leaders routinely apply expert analysis and best practices to situations where cause and effect cannot be known in advance — Cynefin's Complicated-domain tools applied to Complex-domain problems. The result is expensive, expert-driven initiatives that produce no movement because the problem only reveals its nature through experimentation, not analysis. Simultaneously, over-familiarity with Clear-domain routines creates complacency that can cause sudden, catastrophic collapse into Chaos — what Snowden calls 'falling off the cliff' — because no one questioned whether the situation had quietly shifted domains.

The solution

Classify your problem before choosing your strategy. Cynefin ensures you apply the response pattern that fits the actual nature of your situation — not the one you're most comfortable with or most recently trained in.

You bring
  • A specific situation, decision, or challenge you are currently navigating
  • Context about what you already know — or don't know — about cause and effect in this situation
  • Your current response approach, so we can assess whether it fits the actual domain
  • Relevant constraints, stakeholders, or time pressures shaping the decision
You get
  • A Cynefin domain classification for your specific situation, with supporting rationale and key assumptions
  • A domain-appropriate response strategy with concrete actions to take — and explicit anti-patterns to avoid
  • Misclassification warnings if your current approach mismatches the domain you're actually in
  • For Complex situations: a safe-to-fail probe portfolio with defined sensing criteria and success signals for each experiment
In action

Watch the methodology work.

Three specimens from a single real session: the same situation, unaided and calibrated, the full transcript, and the skill answering live in the channel where the work happens.

Fig.01The same situation, two ways
Specimens A · B
Specimen AVague
Unaided

“Your leadership team has spent two months debating a major strategic decision. Half the team wants to apply the approach that worked at their last company; the other half is pulling in consultants to run deeper market analysis. The debate is sophisticated but unresolved, pressure to commit is mounting, and everyone is treating this as a problem with a right answer that more analysis will eventually reveal.”

×Vague — nothing concrete to act on.
Specimen BCalibrated
With Cynefin Framework

“You've classified the situation as Complex — cause and effect are genuinely unpredictable in advance, regardless of how much expert analysis you commission. Instead of committing to one path after more debate, you've designed three parallel safe-to-fail probes, each bounded enough to survive failure, each with sensing criteria defined before launch. The strategy debate is over. The learning has begun.”

Specific, grounded, and ends with one clear next step.
Fig.02A real session, unedited
In-app · assistant
Cynefin FrameworkClaude
Cynefin Framework
Let's use Cynefin to figure out what kind of situation you're actually in — because the right response depends entirely on that classification, and the most expensive mistakes happen when leaders misread the domain. Tell me about the situation you're navigating. What's the core decision or challenge on the table?
We're a mid-size software company and we just lost our biggest customer — 40% of revenue. Leadership is split on whether to double down on our enterprise sales motion or pivot toward a product-led growth model. We've been debating for two months and can't decide.
Two months of expert debate with no resolution is itself a diagnostic signal worth examining. Let me ask the core Cynefin classification question: in your market, how clear is the relationship between cause and effect on this specific choice? When other companies in your space have made the enterprise-vs-PLG call, has there been a predictable pattern of outcomes — or does it vary widely by context, timing, and factors that are hard to identify in advance?
Honestly, it varies a lot. We've seen competitors go PLG and succeed, and others go PLG and fail completely. Same with doubling down on enterprise. There's no clear rule.
That's a critical data point. Outcomes that vary unpredictably even when experts study them is strong evidence this is a Complex situation — not a Complicated one. Here's why the distinction matters: if this were Complicated, the right move would be to assemble the best consultants, analyze the data exhaustively, and derive the optimal answer. Complicated problems have discoverable right answers. But you're describing a situation where cause and effect genuinely can't be predicted in advance, even with perfect information. In Complex territory, the Cynefin response pattern is Probe-Sense-Respond — not Sense-Analyze-Respond. Two months of debate is classic analysis paralysis applied to a domain that resists analysis. You're not going to think your way to the answer. You need to design experiments that let the answer emerge from reality.
That's a reframe I hadn't considered. So we shouldn't be trying to pick the 'right' strategy — we should be running probes? What would safe-to-fail actually look like here?
Exactly. But safe-to-fail probes in the Cynefin sense have specific criteria: each experiment must be genuinely bounded — failure is informative, not existential — you run multiple probes simultaneously in different directions, and you define sensing criteria before you launch, so you recognize signals rather than rationalize results after the fact. For your situation: one probe might be a 90-day freemium tier targeting a specific segment, with defined in-advance metrics — trial-to-paid conversion, activation depth within 30 days, support burden per user. A parallel probe could be two dedicated enterprise hunters in a new vertical, with their own clear signals for traction. Neither is a full bet on a direction. Both are structured experiments in a genuinely uncertain environment. What resources could you commit to a 90-day probe without it being existential if it fails?
We could probably put 2-3 engineers and a small runway slice behind each without it threatening the company. I hadn't thought of running both simultaneously — we kept framing it as a binary choice we had to get right.
That binary framing is the Complicated-domain assumption sneaking into a Complex situation. In Complicated contexts, you pick the best option after analysis. In Complex contexts, you run parallel probes and let the environment tell you what works. The two-month debate is over; you now have two experiments and a 90-day sensing window. One more thing to lock in before you launch: what would a success signal look like for each probe — not a vanity metric, but the leading indicator that would genuinely give you confidence this path can scale?
Reply to Cynefin Framework
What the skill is doing
01Cynefin Domain Classification
02Response Pattern Mapping
03Misclassification Detection
Fig.03The same skill, in the channel you already use
WhatsApp · live
9:41
CFCynefin Frameworktyping…
Messages are end-to-end encrypted. No one outside of this chat can read them.
Today
We're a mid-size software company and we just lost our biggest customer — 40% of revenue. Leadership is split on whether to double down on our enterpr…9:14✓✓
Tell me about the situation you're navigating.9:14
What's the core decision or challenge on the table?9:15
Message

The same skill, where the work happens.

No new app to learn. The methodology runs over the WhatsApp Business API, so the answer lands as a reply in the thread you’re already in — same rigour, zero context-switch.

Reads the situation, names the pattern, returns one concrete next move.
Delivered in seconds, inside a conversation that already exists.
Specimen · WhatsApp Business API · live
Capabilities

What it does, specifically.

Each capability is a distinct move drawn straight from the source methodology — not a generic assistant guessing.

CapabilityC-01

Cynefin Domain Classification

Through structured probing questions about cause-effect visibility, expert predictability, and past intervention outcomes, the skill maps your situation to one of the five Cynefin domains. This classification is the foundational prerequisite — every subsequent response strategy depends on getting it right, and the questions often surface hidden assumptions about the problem.

Based on Snowden's core diagnostic criterion: how discernible is the relationship between cause and effect, and by whom? Can it be identified in advance (Clear), discovered by experts after analysis (Complicated), only perceived in retrospect (Complex), or not discerned at all (Chaotic)?
CapabilityC-02

Response Pattern Mapping

Each Cynefin domain has a distinct and non-interchangeable response pattern. The skill maps your classified situation to the correct pattern — Sense-Categorize-Respond for Clear, Sense-Analyze-Respond for Complicated, Probe-Sense-Respond for Complex, Act-Sense-Respond for Chaotic — and translates each into concrete leadership behaviors, appropriate practice types, and specific next actions.

Directly implements Snowden & Boone's four response patterns as defined in the 2007 HBR article — each calibrated to the epistemic conditions of its domain, and especially the critical distinction between expert authority (Complicated) and distributed emergence (Complex).
CapabilityC-03

Misclassification Detection

The skill actively surfaces signals that your current approach is mismatched with the actual domain — especially the Complicated/Complex confusion that drives most organizational dysfunction. It explains the specific cost of that mismatch in your context and the corrective actions needed to recalibrate.

Addresses what Snowden identifies as the primary failure mode in organizational decision-making: applying Complicated-domain tools (expert analysis, predictive modeling, best-practice replication) to Complex situations where emergent behavior cannot be predicted by any level of expertise.
CapabilityC-04

Safe-to-Fail Probe Design

For Complex-domain situations, the skill helps design multiple small, parallel, reversible experiments — each testing a different hypothesis about which patterns are emerging. It specifies what signals to watch for, what to amplify if a probe succeeds, and what to dampen if it fails, so you recognize emergence rather than rationalize results after the fact.

Implements Snowden's Probe-Sense-Respond pattern for Complex situations. Safe-to-fail probes — explicitly contrasted with 'fail-safe' engineering solutions — are intentionally small and diverse, designed to generate signal about emergence rather than control outcomes.
CapabilityC-05

Cliff-Risk Assessment

For situations classified in the Clear domain, the skill evaluates whether accumulated complacency or unchallenged routine has created vulnerability to sudden, catastrophic collapse. It surfaces warning signs of the Clear-to-Chaos cliff transition before the fall happens.

Based on Snowden's asymmetric domain model: the boundary between Clear and Chaotic is a cliff, not a slope — transitions are sudden and recovery is slow. This makes complacency detection in the Clear domain one of the highest-leverage early-warning functions in the entire framework.
CapabilityC-06

Domain-Shift Recognition

Situations migrate between domains over time: a Chaotic crisis stabilizes into Complex, then Complicated. A formerly Clear process becomes Complex as market conditions change. The skill helps you recognize leading indicators that your situation has shifted domains so your response strategy can be recalibrated before the mismatch becomes costly.

Based on Snowden's concept of dynamic domain transitions: classification is not a one-time event but an ongoing sense-making practice, with liminal zones between domains requiring special attention as situations evolve.
Tested

Graded before it shipped.

Every skill is scored against independent scenarios for methodology fidelity before it goes live — not vibes, a rubric.

What it produces
OutputD-01

Cynefin Domain Assessment

A structured classification of your situation into one of the five Cynefin domains, with explicit cause-effect evidence, confidence level, key assumptions that could shift the classification, and the most dangerous misclassification to guard against.

OutputD-02

Domain-Fitted Response Plan

A concrete action set tailored to your classified domain — specific leadership behaviors, appropriate practice types (best/good/emergent/novel), anti-patterns to avoid, and for Complex situations, a safe-to-fail probe portfolio with sensing criteria defined for each experiment.

OutputD-03

Misclassification Risk Brief

A prioritized set of warning signals that your situation may be in a different domain than your current approach assumes, the specific organizational cost of that mismatch, and the precise corrective actions to recalibrate your response.

The source

Grounded in the original work.

Every answer traces back to a real source and the practitioner who wrote it — not a secondhand summary. Here is the source of record.

Source authorA-01

Dave Snowden

Dave Snowden created the Cynefin Framework while at IBM Global Services in 1999 and is the founder and Chief Scientific Officer of the Cynefin Centre. His 2007 Harvard Business Review article 'A Leader's Framework for Decision Making' (co-authored with Mary Boone) became one of the most widely reprinted articles in HBR's history. Snowden has advised NATO, major consulting firms, government bodies, and healthcare systems on complexity-based decision-making, and the framework underpins formal practitioner certification programs and a 2020 edited volume with contributions from global scholars and practitioners.

Status · Inspired by Dave Snowden’s work — not yet claimed. Are you Dave Snowden?
Primary sourceS-01

A Leader's Framework for Decision Making (Harvard Business Review, 2007); Cynefin: Weaving Sense-Making into the Fabric of Our World (2020)

by Dave Snowden

Founder, Cynefin Centre; Cynefin Framework creator (IBM, 1999); co-author of one of HBR's most reprinted articles; advisor to NATO, McKinsey, Deloitte, and government bodies; practitioner certification program with global reach.

Read the original ↗
Citationcognitive-edge.com
In the build queue

Be first to run it.

Cynefin Framework is being built right now. Leave your email and we’ll tell you the moment it goes live.

Notify meEmail
At launchI have a decision my team has been stuck on for weeks and I want to use Cynefin to figure out what kind of problem it actually is before we keep debating the answer. Can you walk me through the classification process?