Harnessing the Cynefin Framework in the Data Age
There is a lot of hype around how data and technology can help humanity solve all kinds of problems. While the increase volume of data in the world when combined with the technologies that allow us to better leverage the data has allowed humanity to better understand and solve many new problems around the world, data and technology cannot and will not solve all our problems.
Data and technology are convenient tools the serve a purpose when understanding or solving problems; but sometimes not in the way we would think. Understanding or classifying the nature of the problem could be tremendously helpful in better understanding how data could be leveraged.
Introducing the Cynefine Framework
The Cynefin framework, developed by David Snowden at IBM Global Services in 1999, is a sense making tool to help leaders understand the complexity of the problem and identify a suitable approach to the problem. And since then, the framework has been reinterpreted and adapted to suit the challenges of the time; in our case, in the age of data.
The Cynefin classifies problems into 2 main categories - predictable vs unpredictable problems.
Predictable problems
Refer to problems whereby new insights discovered may be deployed to reliably solve problems in the future. For example, baking a cake or sending a rocket to the moon can be classified as predictable problems, once you figured out how to do it reliably once, if you were to repeat the process there is a high chance of success.
Predictable problems may be further classified into obvious and complicated problems.
Obvious (or Simple): In this domain, cause-and-effect relationships are clear, and the correct answers are self-evident. Decisions are straightforward and based on established facts or best practices.
Complicated: While still predictable, this domain requires analysis or expertise as cause-and-effect relationships are not immediately apparent but can be discovered. Good practices are often used here.
Unpredictable problems
Refer to problems whereby data insights discovered in one instance may not be applicable in the next instance. Raising a child or changing an organization culture are classified as unpredictable; as any parent or HR professional knows, what worked before doesn’t always work again.
Unpredictable may be further classified into Complex and Chaotic problems.
Complex: In this domain, cause and effect can only be perceived in hindsight. There are no right answers, and innovative solutions may emerge from the interaction of different agents and forces. Emergent practices are appropriate.
Chaos: This domain is characterized by a lack of clear cause-and-effect relationships. Quick, decisive action is necessary to establish control and stability. Novel practices are often required.
There is a final category of problems – Disorder. This fifth domain is used when it is unclear which of the other four contexts apply. It acts as a state of not knowing what type of causality exists; and prompts decision makers to pause and categorize before responding.
The value of the framework is best illustrated by understanding the things that could go wrong when the approach does not match the nature of the problem.
Applying predictable solutions to unpredictable problems
Oversimplifying complex problems: If leaders treat complex problems (which require adaptive and emergent responses) as if they are merely complicated (solvable through expert analysis), they might overlook underlying patterns and emerging opportunities. This could result in solutions that do not address the root causes, leading to recurring issues or missed innovations.
Rigid planning in Fluid Situations: In chaotic contexts, where rapid action is needed to re-establish order, applying a methodical analysis (suitable for complicated contexts) can waste valuable time and resources, possibly leading to greater disorder or crisis escalation.
Apply Unpredictable solutions to Predictable Problems
Over-complicating simple tasks: Using a probe-sense-respond approach (suitable for complex environments) for simple problems can lead to unnecessary delays, increased costs, and confusion. For example, instead of applying straightforward best practices, an organization might waste time experimenting with new methods for a problem that has a well-known solution.
Unnecessary experimentation in Complicated situations: When a situation requires expert analysis (complicated), introducing multiple simultaneous experiments (a strategy better suited for complex problems) could scatter resources and focus, leading to suboptimal outcomes despite the availability of expert solutions.
Cynefin Framework in the Data Age
When organizations or leaders lean too much on one decision making approach, the chance of being correct in each instance is simply up to chance. The Cynefin framework provides leaders and organizations with the necessary vocabulary to classify problems first before committing resources to decisions.
Recent developments within the realm of data and technology have led to a proliferation of technical terms, approaches, tools and platforms which further complicate matters.
In the context of the data age, consider the following approaches when deploying data and tech to solve problems:
Obvious problems – Suggest that the problem is already well understood with a lot of supporting data that points to the right solution that works each and every time. Most organizations would have Standard Operating Procedures (SOP) built around it. In the data age, such problems are primed automation, where machines can consistently apply the SOPs at scale with no human intervention. The only thing to look out for is when the underlying assumptions for automations change in the future.
Complicated problems – Data Science or Analytics really shines here, because any insights gained can be reliably applied to solve problems in the future. Granted some problems may be outside the wheelhouse for the organizations, employ external help where needed.
Complex problems – We are seeing an increasing number of problems fall into this category. Data can be useful in this category as well, not for solutions but for inspiration. It can provide hints to where the organization might want to explore, understanding that it may not lead to anything concrete. The closest analogy is to treat such problems like an R&D project; something that takes a longer term approach and may not produce results.
Chaotic problems – Data will not help problems in this category. The response remains the same as before, leadership should act first to stabilize the situation and analyze later if applicable.
As the world is changing faster than ever, it is more and more important today for organizations to classify their problems first before making decisions. While data cannot solve all problems, it can help organizations better under the underlying issues or ask better questions for the problems where data can help.
If you’re a leader or part of an organization looking to refine your problem classification and solution strategies for the data age, don't navigate these complexities alone. Contact FYT Consulting today to discover how our expertise with the Cynefin framework can transform your approach to problem-solving, ensuring that you apply the right solutions to the right problems, every time.
Comments