Cynefin Framework

The Cynefin Framework helps you work out what kind of situation you're actually in, before you decide how to respond. Most decisions don't go wrong for lack of effort - they go wrong when a sensible approach is aimed at the wrong kind of problem. Cynefin gives you a way to tell the kinds apart, so your response fits the situation in front of you.

What is the Cynefin Framework?

The Cynefin Framework is a sense-making model created by Dave Snowden in the early 2000s, drawing on complexity science and knowledge management. Its central idea is simple: the right way to make a decision depends on how knowable the cause and effect is. A situation where you can see exactly what leads to what needs a very different approach from one where the connections only become clear in hindsight.

The name is the giveaway. "Cynefin" (pronounced "kuh-nev-in") is a Welsh word meaning habitat, or the place that shapes you. That is why it helps to picture the framework as a landscape rather than a grid of boxes - a stretch of ground with firm, settled high points and shifting, uncertain lower ground, where the terrain you are standing on tells you how to move.

Snowden and Mary Boone introduced it to a wide audience in their 2007 Harvard Business Review article, A Leader's Framework for Decision Making. Their argument was that the familiar tools of management assume clear cause and effect - and quietly fall apart when that assumption doesn't hold. Cynefin sorts situations into five domains: four where you can act, and one in the middle where you can't yet tell which of the four you're in. You can read more on the official Cynefin site.

How the Cynefin Framework works

Think of the four domains as terrain. Height stands for order - how reliably you can trace cause and effect. The high, settled ground is where things are knowable; the low, shifting ground is where they aren't.

That gives you two halves. The ordered domains, Clear and Complicated, are places where cause and effect can be worked out in advance - by anyone, or by an expert. The unordered domains, Complex and Chaotic, are places where they can't; you only see the links by acting and watching what happens. Each domain comes with a matching way of acting and a matching kind of practice. Get the domain right and the right response tends to follow.

Clear - the known knowns

In the Clear domain, cause and effect are obvious to everyone. There is a right answer and it's known. The approach is to sense, categorise, respond: see what's in front of you, match it to a known category, and apply the established procedure. This is the home of best practice - tested, settled steps that work because the situation is genuinely the same every time. Processing an expenses claim, resetting a password, following a safety checklist: you don't need to experiment, you need to follow the rule.

The guiding question is simply: what's the established rule here?

Complicated - the known unknowns

In the Complicated domain, cause and effect still hold - but you can't see them at a glance. There may be several good answers, and finding them takes analysis or expertise. The approach is sense, analyse, respond: gather what you know, work it through (or bring in someone who can), and choose a sound option. This is the realm of good practice rather than a single best one. Engineering a bridge, diagnosing a recurring fault, designing a pay structure - knowable, but only to those who do the work of knowing.

The guiding question is: what does the analysis tell us?

Complex - the unknown unknowns

In the Complex domain, the ground shifts as you walk on it. Cause and effect are real but only visible in retrospect - the same action can produce different results because the situation responds to what you do. Analysis won't get you there, because there's nothing stable yet to analyse. The approach is probe, sense, respond: run small, safe-to-fail experiments, watch closely what they tell you, and amplify what works while damping what doesn't. Solutions emerge rather than arrive. Most questions about people, culture and change live here - which is why the Complex domain overlaps so closely with wicked problems, the kind that can't be solved once and stay solved.

The guiding question is: what can we safely try?

Chaotic - the unknowable

In the Chaotic domain, there are no stable connections to find and no time to look for them. Things are moving too fast, and the priority is to stop the bleeding. The approach is act, sense, respond: do something to establish a bit of order, see whether it holds, and steer from there. Whatever you do will be novel, because the situation is novel. A crisis, an outage, an accident in progress - you act first and understand later. The OODA loop is a useful companion once you're here, giving you a rapid cycle for deciding and acting under pressure.

The guiding question is the bluntest one: what stops the bleeding?

Disorder - not knowing which

The fog in the middle is Disorder: the state of not knowing which domain you're in. It isn't a place you choose to act from - it's where everyone starts, before they've read the situation. The danger of Disorder is that people fall back on whichever domain they're most comfortable in. An expert treats everything as Complicated and reaches for analysis; a firefighter treats everything as Chaotic and reaches for action. The way out is to break the situation into parts and move each part into a domain you can actually name. Often a single tangle is really two or three different problems wearing one label.

The cliff: why the simplest situations carry the biggest risk

The boundaries between most domains are slopes. You drift from Complicated into Complex as things get less knowable, and you can feel the ground softening as you go. But one boundary isn't a slope at all - the line between Clear and Chaotic is a cliff.

Here's how the fall happens. A situation sits comfortably in the Clear domain, so the rule gets applied, the box gets ticked, and attention drifts elsewhere. Nobody's watching the ground any more. Then the conditions quietly change - a rule goes out of date, an assumption stops being true - and because no one was looking, there's no gentle warning. You don't slide into Complicated. You go straight off the edge into Chaotic, and now you're acting in a crisis you could have seen coming.

This is the model's sharpest point, and it's easy to miss: the situations that feel most under control are the ones where complacency costs the most. Treating something as Clear isn't free - it carries a duty to keep checking that it still is.

How to use the Cynefin Framework

Cynefin works as a structured conversation, not a calculation. The point is to slow down for a moment and sort before you solve.

Step 1: Name the situation plainly. Write down the specific decision or problem you're facing. Keep it concrete - "our new starters are struggling in their first month" gives you more to work with than "onboarding is a problem".

Step 2: Decide whether it's ordered or unordered. The fastest test is one question: do we genuinely understand cause and effect here, or are we guessing? If you can trace what leads to what, you're in ordered ground (Clear or Complicated). If you can't - if the situation keeps surprising you - you're in unordered ground (Complex or Chaotic).

Step 3: Place it in a domain and match your response. Once you know the half, the domain usually follows, and the domain tells you how to act - follow the rule, bring in analysis, run an experiment, or act to stabilise. The mismatch to watch for is treating a Complex problem as Complicated: reaching for an expert plan when the situation needs small experiments instead.

Step 4: Watch the boundaries, especially the cliff. Situations move between domains over time. The one to keep an eye on is anything you've filed under Clear and stopped thinking about - that's where the cliff is.

Run this with the group rather than alone. In a session, draw the landscape, take a handful of live issues, and place each one together. The arguments about where something belongs - "I think this is Complex, not Complicated" - are usually where the real understanding happens, because they force people to say out loud how well they actually understand the problem. For leading through the genuinely unordered domains, our guide to adaptive leadership picks up where Cynefin leaves off.

Example

Imagine a leadership team watching satisfaction scores fall in one of their services. Their instinct is to treat it as Complicated: commission a review, have experts analyse the data, and produce a plan. The plan is thorough and well argued - and three months on, the scores haven't moved.

The reason is that the problem was never Complicated. It was Complex. The dip came from a shifting tangle of new staff, raised expectations, a quiet change in team culture and a process tweak that interacted in ways nobody could have predicted. There was no stable cause to analyse, so the expert plan was answering a question the situation hadn't asked.

Working in the Complex domain, the team changes tack. Instead of one big plan, they run several small probes at once - a different shift pattern in one area, a change to how new starters are paired, a weekly check-in trialled in a single team. Some help, some do nothing, one makes things briefly worse. They keep the ones that work and drop the rest. Over a few cycles, an approach emerges that no plan could have specified up front, because it had to be discovered in contact with the real situation.

The framework didn't tell them the answer. It told them what kind of problem they had - and therefore that probing, not planning, was the way through.

Limitations

It helps you think; it doesn't decide for you. Cynefin is a sense-making aid, not an algorithm. It clarifies what kind of situation you're in and what sort of response fits, but it won't hand you the decision. The work of choosing still belongs to you.

The boundaries are genuinely fuzzy. Real situations rarely sit neatly inside one domain, and sorting can be a judgement call. That's a feature, not a flaw - but it means two sensible people can place the same problem differently, and the framework won't settle the argument for them.

The labels can be misused. "It's complex" can become a way to avoid accountability or dodge a decision that's actually fairly Clear. Used loosely, the language gives cover for inaction. The discipline is to sort honestly, not conveniently.

Situations don't stay put. A domain describes where something is now, not where it will be. Problems migrate - especially over that cliff - so a placement made once needs revisiting rather than treating as settled.

Getting started

Take one decision your team is wrestling with this week and ask the single sorting question: do we genuinely understand cause and effect here, or are we guessing? That one answer tells you whether to reach for analysis or for experiments, and it's often enough to break a stuck conversation - because half the disagreement in a room usually comes from people unknowingly arguing from different domains.

From there, the framework pairs well with a wider view of complexity. The Iceberg Model helps you explore the deeper structures and beliefs that create the complexity Cynefin asks you to navigate, and VUCA Prime offers a leadership response to the volatility and uncertainty the unordered domains throw up. If you want to build this kind of thinking across a team, our systems thinking course and our work on strategic alignment both start from the same principle: read the situation honestly before you commit to a response. Cynefin also connects to Tuned to Change, the dimension of our EMERGENT framework concerned with how readily an organisation senses and responds to shifting conditions.