The Cynefin Framework: Matching Your Approach to the Situation

Most leaders have a favorite way of deciding. Some gather data and analyze; some call in experts; some trust their gut and act. Each approach works well in some situations and badly in others. The trouble starts when a leader uses the same approach everywhere.

The Cynefin framework, created by Dave Snowden, is a tool for noticing what kind of situation you are in before choosing how to respond. In 2007, Snowden and Mary Boone presented it to Harvard Business Review readers in "A Leader's Framework for Decision Making," a piece the official Cynefin wiki says Snowden has described as "the first touchstone of the third age of Cynefin."

This guide explains the framework's domains in plain terms, what the framework's creators suggest doing in each, and how to apply it to real problems, which rarely sit neatly in one place.

Where the framework comes from

Cynefin, pronounced kuh-nev-in according to The Cynefin Company, is a Welsh word. The company describes it as signifying the multiple, intertwined factors in our environment and experience that influence how we think and act. The official Cynefin wiki traces the framework to the late 1990s, in the fields of knowledge management and organizational design, with the first reference to it as a framework dating to 1999.

The 2007 HBR article opens with a story that captures the problem it addresses: a police deputy chief in Palatine, Illinois, who after a 1993 shooting suddenly had to handle grieving families, a frightened community, a busy department and a flood of media, all at once.

The wiki stresses that Cynefin is "a decision support framework" rather than a method or model. It rests on a principle of "bounded applicability": "there are few, if any, context-free solutions but many valid context-specific ones."

The domains in plain language

According to the Cynefin wiki, the framework starts with three primary kinds of systems, defined by their constraints: ordered, complex and chaotic. Order is then split in two, giving these domains:

• Clear: cause and effect are self-evident to any reasonable person. The wiki calls this the only legitimate home of best practice. Its example: the checklist of items not allowed in your hand luggage on a plane. Earlier versions called this domain Simple, then Obvious; Snowden adopted the name Clear in 2020.
• Complicated: cause and effect exist, but finding them requires expertise or analysis. Good practice applies. Example: the engineering process of building a bridge.
• Complex: many interacting parts, with no straight line from cause to effect. The only way to understand the system is to interact with it. Example: a children's party.
• Chaotic: no effective constraints. Turbulence prevails and immediate stabilizing action is needed.
• Confused: the central area, where you do not know which domain you are in. Earlier versions called it Disorder, and its liminal part is now called aporetic. The wiki notes it is frequently, but wrongly, confused with chaos.

The wiki also warns against drawing these as a two-by-two grid. The domains are not quadrants, and their arrangement is meant to show how situations move from one to another.

What to do in each domain

In a March 2022 post on The Cynefin Company's blog, Snowden summarized where to focus in each domain:

• Clear: apply best practice and focus on compliance with the rules, but also watch for workarounds and noncompliance. Those, he wrote, are early weak signals that you are over-constrained and at risk of a collapse into chaos.
• Complicated: the central question is how much you trust the expertise you are relying on, since the domain depends on people who know the answer.
• Complex: identify every coherent hypothesis about what might improve things and "probe in parallel on all coherent hypotheses not do them in sequence." Probes should be safe-to-fail, which, he notes, "doesn't mean you won't fail, it means if you do fail then you can recover." Then watch for emerging patterns and amplify or dampen accordingly.
• Chaotic: "Make decisions fast to keep your options open." Stop the leak, aim for a temporary stabilization that lets you survive, then work on more resilient solutions.
• Confused, or aporetic: Snowden suggests this is "probably where you should start," by breaking the issue down before deciding which parts belong where.

Break big problems into parts

Real problems usually span several domains. Snowden's advice is to decompose a situation to its "lowest level of coherent granularity," the smallest pieces that still make sense on their own, and then treat each piece on its own terms.

For example, imagine a small bakery chain whose main flour supplier suddenly stops shipping after a plant fire:

• This week's shortage is chaotic. Act now: buy at spot prices, cut low-margin items and tell stores what to prioritize.
• Payroll, food-safety rules and allergen labeling stay clear. Follow the established procedures without improvising.
• Adjusting recipes for a different flour is complicated. Bring in someone with baking science expertise.
• How regular customers will react to slightly different bread is complex. Run small, parallel tests in a couple of stores and watch what happens before changing every location.

One crisis, four different kinds of decision. Treating all of it as chaos leads to panic; treating all of it as clear leads to rigid rule-following when rules no longer fit.

Watch for movement and the cliff

Situations shift between domains. The wiki describes the most stable pattern as a constant back-and-forth between complex and complicated, with some material moved into clear once it is stable enough and the shift is low risk.

It also describes a danger zone. The boundary between clear and chaotic is drawn as a cliff, where it is "all too easy to walk blindly off the cliff through excessive confidence in the applicability of rigid constraints." Organizations that treat a changing environment as routine can fall from order straight into crisis.

A field guide produced by the European Commission's Joint Research Centre with the Cynefin Centre adds crisis advice: in a crisis, decisions must be made quickly and authority and hierarchy are useful. It also recommends setting, early on, the triggers for shifting back to normal operations once the crisis is over.

Common mistakes

• Using the framework as a one-time sorting exercise. Snowden writes that working with Cynefin "is never about static categorisation."
• Calling every hard problem complex as a reason not to analyze. Many problems are merely complicated and need expertise, not experiments.
• Copying best practices from another organization into a complex situation where context changes everything.
• Forgetting the confused middle and assuming you already know which domain you are in.

Key takeaways

• Cynefin helps you identify what kind of situation you face before choosing how to decide.
• Clear situations suit best practice; complicated ones need expertise; complex ones need safe-to-fail experiments; chaotic ones need fast stabilizing action.
• Break large problems into parts, since one crisis can contain several domains.
• Watch for workarounds and complacency in routine operations; they can signal a coming fall into chaos.
• In a crisis, set triggers for returning to normal operations.

Sources

• The Cynefin Company, About the Cynefin Framework
• Cynefin.io (official wiki), Cynefin (last modified April 16, 2023)
• Cynefin.io (official wiki), Cynefin Domains (last modified August 3, 2022)
• Dave Snowden, Cynefin St David's Day 2020 (1 of 5), Cognitive Edge blog (March 1, 2020)
• Dave Snowden, Cynefin St David's 2022 (1 of 2), The Cynefin Company blog (March 1, 2022)
• Cynefin.io, Field Guide to Managing Complexity (and Chaos) in Times of Crisis
• David J. Snowden and Mary E. Boone, A Leader's Framework for Decision Making, Harvard Business Review (November 2007)