The cynefin framework: choosing an approach when the answer is unclear

What is the cynefin framework?

The Cynefin framework is a sense-making framework that helps you choose how to respond to different kinds of situation. Its central idea is that different problems need different approaches. A reliable procedure may solve one problem, expert analysis another, while a third requires experimentation rather than a predetermined answer.

Cynefin distinguishes four main domains:

Clear: cause and effect are apparent, so use established practice.
Complicated: cause and effect can be understood through investigation, so seek relevant expertise.
Complex: outcomes emerge through interactions and cannot be reliably predicted, so use small, bounded experiments.
Chaotic: the situation is too unstable to establish useful cause-and-effect relationships before responding, so act to stabilise it.

There is also a central state of confusion, where you have not yet established which approach fits.

For leaders in further education, skills and workplaces, the practical question is:

What kind of situation are we facing, and what does that mean for our next action?

Cynefin does not make the decision for you. It helps you avoid applying the wrong kind of decision-making.

Where cynefin comes from

Dave Snowden developed Cynefin from the late 1990s, including through his work at IBM. The Welsh word cynefin evokes habitat and the accumulated relationships and experiences that shape our circumstances.

Two important publications developed and explained the framework:

Cynthia F. Kurtz and David J. Snowden’s 2003 paper, The new dynamics of strategy: Sense-making in a complex and complicated world.
David J. Snowden and Mary E. Boone’s 2007 Harvard Business Review article, A leader’s framework for decision making.

The terminology has evolved. Older explanations use “simple” or “obvious” for what is now usually called clear, and “disorder” for the central area now commonly called confusion. More recent treatments further distinguish forms of confusion.

Cynefin is an influential practitioner framework grounded in ideas from complexity and sense-making. It is not a validated diagnostic test that objectively assigns every problem to a domain. Commercial consultancy and training are offered around it; those services should not be confused with independent evidence of effectiveness.

Why it matters

Leaders often have a preferred response. Some write a procedure. Some commission analysis. Others organise a workshop or make a rapid decision.

Each response can help, but only in suitable circumstances.

For example:

Adding procedures may not resolve low learner participation if the causes differ between groups and change over time.
Running experiments is inappropriate when an established emergency procedure requires immediate action.
Treating a technical calculation as a matter of opinion wastes expertise.
Demanding a guaranteed result from a new partnership may conceal genuine uncertainty.

Cynefin encourages a better match between the situation and the response. It also provides language for constructive disagreement. Instead of arguing about whether someone is being indecisive or controlling, colleagues can ask whether they understand the nature of the problem differently.

Understanding the domains

The framework’s response sequences describe the logic of each approach, not a rigid script.

Domain	What you can reasonably know	Response sequence	Appropriate starting point
Clear	Cause and effect are apparent and repeatable	Sense, categorise, respond	Apply an established procedure
Complicated	Cause and effect can be established through expertise or analysis	Sense, analyse, respond	Investigate and compare options
Complex	Patterns become understandable through interaction and observation, often retrospectively	Probe, sense, respond	Run bounded experiments and learn
Chaotic	Useful patterns cannot yet guide an immediate response	Act, sense, respond	Contain harm and establish stability
Confusion	You do not yet know which domain applies	Separate and investigate	Clarify what you are dealing with

Clear: use established practice

A clear situation has a dependable relationship between action and outcome. People can recognise the circumstances and apply an appropriate rule.

For example, a correctly completed expenses claim falls within an agreed policy. The administrator checks it against the criteria and processes it through the established workflow.

The leadership task is to ensure that:

The procedure is accessible and understood.
Staff have the resources and authority to follow it.
Exceptions can be raised rather than forced through the process.
Checks identify when the procedure no longer fits.

Cynefin associates this domain with best practice, but that does not mean one procedure is universally best. It means a reliable approach exists for the defined circumstances.

The danger is complacency. Repeatedly treating unusual cases as routine can hide changes or accumulating failures.

Complicated: bring in expertise

In a complicated situation, an answer can be found, but it is not immediately obvious. Investigation, specialist knowledge or comparison of alternatives is needed.

Imagine a college’s payroll system produces an unexpected pension deduction. A payroll specialist may need to examine the employee’s circumstances, relevant rules, configuration and transaction history.

There may be several defensible solutions. Cynefin therefore associates this domain with good practice, rather than assuming a single best answer.

Useful leadership questions include:

What expertise do we need?
What evidence would distinguish the possible explanations?
What assumptions are the specialists making?
When will further analysis stop adding enough value?

Complicated does not simply mean “difficult”. An intricate engineering calculation may be complicated because established methods can resolve it. A seemingly simple question about changing team behaviour may be complex.

Complex: learn through bounded experiments

In a complex situation, many interacting influences shape what happens. People respond to one another, adapt to interventions and change the conditions.

You may explain an outcome afterwards without being able to predict it reliably beforehand. A convincing retrospective story is not necessarily a dependable recipe for next time.

Consider an employer trying to improve collaboration between departments. Reporting lines matter, but so do trust, incentives, workload, informal relationships and previous experiences. Copying another organisation’s meeting structure may produce different results.

The response is probe, sense, respond:

Try small interventions with explicit boundaries.
Observe what happens, including unexpected effects.
Expand promising patterns and stop or adapt unhelpful ones.

These are often called safe-to-fail probes. “Safe to fail” means that an unsuccessful intervention has tolerable, contained consequences. It does not mean accepting harm to learners, staff or essential services.

Where feasible, several small, different probes provide more learning than one large pilot tied to a favoured solution.

Chaotic: stabilise before analysing fully

In a chaotic situation, instability prevents useful analysis from guiding the immediate response. The first priority is to establish enough order to make further decisions possible.

For example, a sudden building incident during a working day may involve incomplete information, disrupted communication and immediate risks.

The response is act, sense, respond:

Take proportionate action to contain harm.
Check whether the action is improving the situation.
Adjust and establish a more manageable environment.

This does not mean improvising instead of following emergency procedures. Existing plans, trained roles and clear responsibilities provide essential support.

Nor does it mean maintaining command-and-control indefinitely. Once the immediate situation is stable, technical investigation and collaborative learning may become appropriate.

Confusion: find out what you are dealing with

Confusion is not a fifth problem-solving method. It describes not knowing which domain applies.

Different people may default to their preferred approach: an analyst asks for more data, an operational manager asks for a procedure, and a senior leader demands immediate action.

The useful response is to break the situation into smaller parts.

A new apprenticeship programme might contain:

Clear administrative requirements.
Complicated scheduling and resource calculations.
Complex employer and learner engagement.
A chaotic element if a critical delivery arrangement suddenly collapses.

Trying to label the whole programme with one domain would hide these differences.

Applying cynefin in an FE setting

Example: participation falls on an adult learning programme

A programme team notices that attendance has become less consistent. Its first suggestion is a standard warning email to everyone who misses a session.

Cynefin prompts the team to separate the issues.

Checking whether registers contain errors is largely a clear task, using established checks. Investigating whether timetable clashes affect particular learners may be complicated, requiring analysis.

Understanding why learners disengage is likely to contain complex elements. Work patterns, confidence, transport, caring responsibilities and relationships with tutors may interact.

The team could:

Speak with learners and staff to understand their experiences.
Test two or three small changes, such as a different reminder format or an optional catch-up arrangement.
Define what failure would look like, including additional burden or reduced access.
Review attendance alongside learner feedback and staff workload.
Adapt before expanding.

These are illustrative actions, not evidence that any one intervention will work. Essential support and reasonable adjustments should not be withheld to create a comparison.

Example: a workplace booking system goes offline

A training business loses access to its booking system just before several courses begin.

Different elements need different responses:

Chaotic: establish who is coordinating the response and ensure arriving participants receive consistent instructions.
Clear: activate an agreed continuity procedure where one exists.
Complicated: ask the system provider or technical team to diagnose the fault.
Complex: explore how repeated disruption has affected customer trust and staff confidence.

The lesson is not that every outage is chaotic. A routine outage with a working fallback may be manageable through established procedures. Context determines the approach.

A practical checklist for your next decision

Use Cynefin in a short team discussion before committing to a substantial response.

1. Define the decision

Write a specific question, such as:

“What should we do next to understand and reduce missed workplace assessments?”

Avoid framing the task as confirming a preferred solution.

2. Establish constraints

Identify immediate safety concerns, legal duties, contractual requirements, deadlines and unacceptable consequences.

A complex situation does not suspend your obligations.

3. Compare concrete experiences

Ask colleagues for examples of what has actually happened before choosing a domain.

Useful prompts include:

Can a competent person follow an established procedure here?
Could a relevant specialist establish the answer?
Will people’s responses change what happens next?
Do we need immediate containment before further investigation?
Are we combining several different situations?

4. Match the next action to the domain

Choose a proportionate response:

Clear: use or repair the procedure.
Complicated: obtain targeted analysis or expertise.
Complex: design bounded probes with feedback.
Chaotic: stabilise and reassess.
Confusion: separate the issues and gather enough information to proceed.

5. Set a review point

Record:

What you expect to learn.
What you will observe.
Who can stop or adapt the action.
When you will review it.
What would make you reconsider the domain.

For a complex probe, include an explicit boundary: “We will stop or redesign this if it increases missed learning or creates an unsustainable workload.”

Limitations and common misunderstandings

Cynefin is not a personality model. It describes situations, not types of leader or learner.

Complex does not mean very complicated. The distinction concerns the nature of cause and effect, not the amount of effort required.

Urgency does not automatically mean chaos. A time-sensitive task can still have a known procedure.

Expertise still matters in complex situations. Experts help identify risks and design sensible interventions, but cannot necessarily predict outcomes.

A probe is not proof of causation. Small experiments can inform local decisions, but apparent improvements may have other explanations. Formal impact claims require suitable evaluation methods.

Domain labels are provisional. People can disagree, boundaries can be unclear, and situations can change. The conversation should improve action, not become a classification exercise.

Related approaches

Cynefin helps you decide what kind of approach to use. Other methods help you carry it out:

Process mapping and checklists support reliable work in clear situations.
Technical analysis and structured options appraisal support complicated decisions.
Systems thinking helps reveal relationships, feedback and unintended consequences.
Plan-do-study-act cycles can structure small tests, though using a cycle does not automatically make a test safe or informative.
Emergency and business continuity plans support stabilisation and recovery.

These approaches are complementary rather than interchangeable.

Summary and your next step

Cynefin offers a practical discipline: do not assume every problem needs the same response.

Use established practice when the relationship between action and outcome is clear. Seek expertise when analysis can establish an answer. Experiment carefully when outcomes emerge through interaction. Act to stabilise when conditions are chaotic, and separate the issues when you are confused.

Your next step: take one live team challenge and spend 15 minutes identifying its different elements. Agree one next action, one boundary and one review point.

Sources and further reading

Kurtz and Snowden: The new dynamics of strategy. The 2003 IBM Systems Journal paper provides an early primary account of Cynefin and organisational sense-making.
Snowden and Boone: A leader’s framework for decision making. The 2007 leadership article explains the domains and response sequences. It uses earlier terminology and may require a subscription.
Cynefin wiki: Cynefin. A framework-community resource for terminology, development and more detailed explanations.
The Cynefin Company. A primary practitioner source associated with the framework’s development. It also sells consultancy and training, so its materials should be read with that commercial context in mind.