Before You Solve It, Diagnose It: The Cynefin Framework for Leaders

Picture a fairly typical AI rollout. A management team spends the better part of a year on it: a long vendor bake-off, a training curriculum, a launch date with a return-on-investment target attached. Three months in, usage numbers are flat, and it turns out a handful of people have already moved on to a different tool nobody approved, because it just worked better for what they actually needed. Leadership can't quite explain it. The plan was sound. The execution was sound. And yet.

It's a familiar pattern, and not because anyone involved did anything wrong by their own logic. Most of us default to the same problem-solving posture no matter what we're facing: gather information, build a plan, execute. It's what most management training teaches, and it's genuinely the right approach for a lot of problems. But not all of them. Some situations can't be planned your way through, only learned your way through, and mistaking one for the other is an expensive habit.

The Cynefin framework, developed by Dave Snowden in the late 1990s while he was at IBM, offers a way to avoid this mistake. Cynefin, a Welsh word that roughly translates to habitat or place of belonging, is not a categorization scheme for sorting problems into neat boxes. It is a sense-making tool that helps leaders diagnose what kind of system they are actually operating in before they choose how to respond. That diagnosis, done well, changes everything about the decision that follows.

Four Domains, Four Very Different Responses

Cynefin identifies four primary domains, distinguished by how clearly cause and effect can be understood, plus a fifth condition of disorder that describes not yet knowing which domain applies (Snowden & Boone, 2007).

In the clear domain, sometimes called the obvious domain, cause and effect are well understood and repeatable. Best practices exist and reliably work. The right sequence is to sense the situation, categorize it against a known pattern, and respond according to established procedure (Snowden & Boone, 2007). Processing a routine invoice or onboarding a new hire with a standard checklist both belong here. The risk in this domain is complacency: leaders who assume a situation is obvious can miss the early signals that it has actually shifted into something less predictable.

In the complicated domain, cause and effect still exist and can be known, but discovering the relationship requires expertise or analysis. There is a right answer, or at least a defensible best answer, and finding it rewards skilled judgment. Structuring a financing deal, engineering a bridge, or diagnosing a mechanical failure all live here. The sequence is sense, analyze, and respond (Snowden & Boone, 2007). Most professional training, and most of the consulting industry, is built for this domain, because it rewards rigorous analysis with dependable outcomes.

In the complex domain, cause and effect can only be understood in retrospect. The system involves many interacting agents, often including human behavior, that resist full modeling in advance (Snowden & Boone, 2007). Launching a product into an uncertain market, shifting an organization's culture, or introducing a new technology into a workforce with its own habits and incentives all belong here. The correct sequence is probe, sense, and respond. Leaders run small experiments designed to be safe to fail, observe what actually happens, and let the real feedback shape the next move. Expertise still matters, but it cannot substitute for observation, because the system will not hold still long enough to be fully analyzed before it needs a response.

In the chaotic domain, no discernible relationship between cause and effect exists yet. A system-wide outage or a sudden crisis belongs here. The only responsible sequence is act, sense, and respond: stabilize first, observe the effect of that action, and adjust from there (Snowden & Boone, 2007). Waiting for analysis during genuine chaos is itself a choice, and usually the wrong one.

The fifth condition, disorder, sits at the center of the model and describes not knowing which domain you are actually in (Snowden & Boone, 2007). This may be the most dangerous place a leader can stand, because people in disorder tend to retreat toward whichever domain feels most comfortable to them personally rather than the one the situation demands. An executive who built a career on rigorous analysis will often try to analyze a complex problem into submission. An executive who built a career on decisive action will sometimes act prematurely on something that actually calls for patient experimentation. The framework's most practical use may simply be the discipline of pausing long enough to ask which domain is actually in play.

Why This Matters for AI Adoption

This distinction is especially relevant right now, as organizations make decisions about artificial intelligence. Many leadership teams approach AI adoption as though it were a complicated problem: select the right vendor, follow a structured implementation playbook, train the workforce according to a fixed curriculum, and expect a predictable return. But AI adoption inside a real organization, with its existing culture, its skeptics and its early adopters, its informal workflows and its internal politics, behaves like a complex system rather than a complicated one (Snowden & Boone, 2007). Cause and effect only become visible after the fact.

The organizations that navigate this well tend to treat early AI initiatives as safe-to-fail probes rather than fully specified rollouts (Kurtz & Snowden, 2003). They pilot a tool with a willing team, watch closely for what genuinely changes in how people work, amplify what proves useful, and quietly retire what does not, instead of committing the whole organization to a plan built on assumptions about how the technology will land. This is not a failure of ambition. It is an accurate reading of the terrain.

A Practical Discipline

The habit Cynefin teaches is simple to describe, even if it takes real discipline to practice. Before reaching for a plan, a playbook, or an expert opinion, pause and ask what kind of system is actually in front of you.

If the relationship between action and outcome is well established, follow the best practice and move on. If uncovering that relationship requires expertise, bring in the right analysis. If the outcome can only be understood by observing what happens when you act, run the smallest experiment that will give you real information, and be genuinely willing to let what you learn change your plan. And if the situation is truly chaotic, act first to create stability, and save the explanation for later.

Leaders rarely struggle because they lack a framework; they struggle because they apply the same framework to every situation, regardless of the ground beneath them. The real value of Cynefin is not the map itself. It is the habit of checking the map against the terrain before taking the next step, a discipline Snowden and Boone (2007) describe as central to sound leadership judgment.

References

Kurtz, C. F., & Snowden, D. J. (2003). The new dynamics of strategy: Sense-making in a complex and complicated world. IBM Systems Journal, 42(3), 462-483. https://doi.org/10.1147/sj.423.0462

Snowden, D. J., & Boone, M. E. (2007). A leader's framework for decision making. Harvard Business Review, 85(11), 68-76.

Copyright © 2026 by Severin Sorensen. All rights reserved.