Managing complexity in projects: Extending the Cynefin framework

Recommendations for managing complexity in projects acknowledge different levels of complexity, but often neglect the importance of (non-)repetition. This paper uses the Cynefin framework, structuring problems along the levels of simple, complicated, complex, and chaotic, and the learning loop model as simple but practical tools to sort different project situations. Combining the degree of complexity with repeatability or uniqueness allows a typology of projects to guide management. Single-loop and double-loop learning from past experiences is restricted to the occurrence of similar situations, slicing, project management guidelines and frameworks to complicated projects. The more complex and novel the problem, the more likely that non-standard, non-algorithmic ways such as sensemaking, trial-and-error, and whole-systems approaches will yield satisfactory results. Expertise is useful for simple and complicated projects; however, for new and complex projects, experience, authority and trust, and recognition of abstract patterns (deutero learning) are more important.