Why do we need complex analyzes

Learn more

How can I (best) approach a problem or a question? Is there a perfect solution? Are there maybe several paths leading to the goal? Who could be helpful in finding a solution? Does my "favorite" approach fit the problem or is it the problem itself? With the Cynefin Framework Model, which David Snowden and Mary Boone published for the first time in 2007 in Harvard Business Manager and which Snowden has further developed over the years, answers to these important questions can be found.

Based on Snowden / Boone, the Cynefin Framework Model can be used to assign many issues in the areas of leadership, project management and other fields of work to different system types. Each system is subject to its own logic and functionality, which in turn require different problem-solving approaches. The term system can be understood here very broadly and as a "collection" of several elements that are functionally related. This can include, for example, technical systems such as a machine, systems based on rules / structures such as a business process or social systems such as a team or markets. The Cynefin Framework model helps us to differentiate between four system types in order to be able to draw conclusions for possible approaches on their basis:


Simple systems: Here the connections between cause and effect are obvious and unambiguous. Systems that follow this characteristic can be, for example, a bicycle, a heating circuit or an administrative process. Dealing with a problem follows the process of perceiving - Categorize / assess - react. A problem is identified and categorized so that the best solution can then be applied. The so-called "best solutions" (best practice) to be applied provide good orientation here.


Complicated systems: Here cause and effect are no longer obvious. However, they can be determined with a corresponding effort. A large airliner or a space shuttle can be examples. Most business-oriented management approaches also follow this logic. Therefore, the recommendation for action is perceive - Analyze - react. The best-practice solutions based on the one-best-way approach no longer apply here either. A problem is perceived and this is followed by analyzes, the collection of data and expert opinions, calculations, etc. in order to then uncover the cause-effect relationships and select a solution. Orientation towards good practice examples can be helpful here, which suggests that there can be several "correct" solutions.


Complex systems: The connection between cause and effect remains in the dark in advance. At best, they can be recognized afterwards. For example, all systems of human interaction correspond to this type of system. Orientation towards expert opinions, analyzes, best or good practice is no longer useful here. In order to still be able to act, Snowden / Boone recommend the following sequence: Experiment - Perceive - React. It is important to literally throw a stone into the water and notice what is happening in order to come to solutions that can also be unorthodox. The conscious turning away from the idea of ​​plannability is just as essential for this as opening up to ignorance and an attitude free of value judgments in order to also perceive the emerging solution. Snowden / Boone call this emergent practice. This way of dealing with problems comes close to the systemic counseling attitude, the ideas of Otto Scharmer's U-Theory or the agile arsenal of methods. This logic eludes everyday thinking, since we, as rationally thinking people, are used to thinking and interpreting in linear cause-and-effect relationships.


Chaotic systems: Cause and effect no longer have a relationship here - or they are beyond our understanding; Examples are crises, disasters or emergency situations. In a business context, this type may be found before a bankruptcy or during a turnaround. Snowden / Boone name the novel practice here, which over Act - Perceiving - Reacting is to be tapped. Chaotic systems are in a kind of alarm mode. Immediate action is required here, as there is no time to experiment or analyze. In this case, systems need to be stabilized by acting quickly. Then it is necessary to perceive the (out) effect of this action in order to react or to act again.

The authors describe a fifth domain in the Cynefin Framework model: "Disorder", ie "disorder" or "confusion". This indicates that it is still unknown which type of system the question can actually be assigned to.


Fig .: Cynefin Framework Model (based on: Snowden / Boone)


According to Snowden, the Cynefin Framework Model should be used less for fixed categorization and more for "making sense", ie for orientation and mutual understanding. It is important that the boundaries between the systems are not rigid. The assessment of which type of system is characteristic can also depend on the observer or his level of knowledge, as it can change over time. For example, an initially complex system can become a complicated system through our learning about its functioning and laws.

Use and application

According to Snowden, we are mostly in the "Disorder" area. In hectic everyday life, we therefore do not consciously know which type of system we are dealing with or we lack a general understanding of these differences. In addition, due to the profession we have learned, our work environment and our biography, we tend to prefer one area. In practice, this could be shown in the following "typical answers", for example:

Simple systems:Problems are mostly faulty processes. So we just have to further optimize the processes.

Complicated systems:The problem is rooted in an inadequate database. Therefore, we need time, resources and other experts to solve it.

Complex systems:The problem is overly complex. Therefore we have to bring as many people as possible together and hope that the right solution will emerge.

Chaotic systems:We don't have time to understand the problems and get to the bottom of them. Therefore we need a "strong man" who cleans up and does everything right.


The Cynefin Framework Model makes it clear that the successful problem-solving strategies of one type of system can be less helpful or even counterproductive for another type of system. It enables us to question existing and preferred solution strategies and to find appropriate access. One of the main benefits of this model is to be able to determine, alone or in a team, which type of system can correspond to the question. This can enormously expand our room for maneuver, as despite all professionalism, in everyday life - and especially under pressure - we tend to look at the world and its challenges through a familiar window. On this basis, for example, depending on whether it is complicated or complex, the question can be asked, "Which experts can help us?" or "Which experiment do we want to start with?" In general, this model can support us in making possible fixations of our interpretations and action strategies more variable and thus also richer and more effective.

No system type and the associated recommendations for action are therefore better or worse than the others. The model asks us to look for the purposefulness of our solution strategy. Discussions as to whether it is good or bad to establish an error culture, set up agile teams or ensure strict compliance with rules and standards are becoming less important. "Right and wrong" can become "suitable and unsuitable" in relation to the system type. Solutions can be sought and applied appropriately to the degree of complexity.

Food for thought ...

  • If it is stated in the room that there is only one correct solution, it should be checked whether it is actually a simple system.
  • In which cases is expert advice suitable (e.g. Roland Berger, BCG, etc.), in which cases, for example, is a systemic consultant required and in which cases both or neither?
  • Which system logic do widely used management tools (e.g. quality management, target agreements, competence management, budget planning, etc.) correspond to and to which system types are they used? Do the tools fit or do you need other approaches?
  • In which cases are solutions, tools, processes, measures, etc. that arise from the logic of simple or complicated systems applied to complex systems (such as people, teams, etc.)? At what price? Are there any better alternatives?
  • Do the solutions from the areas of agility, new work, design thinking, etc. fit all operational issues? Where should it be refrained from if necessary?
  • For which system logic is an error culture indispensable and where is it better to avoid it?
  • In which "world" do I move myself and in which my colleagues, employees, partners, superiors or customers? What is the result of this?
  • What does this mean for leadership? And what about designing and managing projects?

At the RKW Competence Center, we try to reflect this diversity in our products:

  • In our business model development process, we offer tools that support the development of expertise for decision-making (complicated) and at the same time offer a circular process design so that you can step back in the process after an "experiment" (complex).
  • The RKW management navigator contains elements of complicated systems (e.g. management tasks) and does justice to the "object" human (complex systems) through the elements of creating relationships or self-management.
  • With 50 examples of good practice (complicated systems), the RKW Digitization Cockpit provides specific orientation and thus helps to expand one's own knowledge base in order to facilitate the (further) development of a digital business model.

author

Sascha Hertling

hertling (at) rkw.de