Framework for Tough Calls: Match Decision to Context

The framework referenced is the Cynefin framework, developed by Dave Snowden at IBM in 1999 as a sense-making device. "Cynefin" (pronounced "kuh-NEV-in") is a Welsh word for "habitat" or "place," reflecting the idea that all individuals have connections that influence their perceptions. The framework aims to provide decision-makers with a "sense of place" to better understand and act on a situation. The model outlines five distinct contexts, or domains, for decisions: clear, complicated, complex, chaotic, and a central domain of confusion or disorder. The domains on the right (clear and complicated) are considered "ordered," where cause and effect are either known or discoverable. The domains on the left (complex and chaotic) are "unordered," where the relationship between cause and effect can only be understood in hindsight, if at all. In the *clear* domain, the relationship between cause and effect is obvious to all; the appropriate response is to "sense, categorize, and respond" by applying best practices. For engineering managers, this applies to routine processes like code check-ins or expense reports. The *complicated* domain involves "known unknowns," where expert diagnosis is required. Here, the approach is to "sense, analyze, and respond," which could involve bringing in senior engineers to investigate a performance issue. The *complex* domain is the realm of "unknown unknowns," where solutions emerge through experimentation. For software development, this mirrors the Agile methodology, using a "probe, sense, respond" approach to iterate on features and get feedback. Many significant business and technology challenges, like developing a novel product, fall into this category. The *chaotic* domain is where cause and effect are unrelated, such as during a major system outage or crisis. The required leadership action is to "act, sense, and respond" to first establish order. This is the only domain where a top-down, directive management style is appropriate to stabilize the situation before moving it into a more manageable context. For an engineering leader, this framework provides a model for executive communication. Instead of presenting every issue in the same way, a manager can frame updates based on the context. For a *complicated* problem, the update would focus on the experts consulted and the data analyzed. For a *complex* problem, the focus would be on the experiments run, the patterns observed, and the next set of probes. This structured approach moves beyond generic advice, offering a concrete way to categorize challenges and articulate the decision-making process to senior leadership. By clarifying whether a situation requires following best practices, engaging experts, running experiments, or taking immediate decisive action, a manager can reduce ambiguity and build trust with their teams and superiors.