Set conditions for insight

• Prioritize transparency over objectivity
• Prioritize understanding over blame or fault
• Resist narrative until late in assessment
• Articulate mental frameworks before debating the resulting positions
• Hold strong opinions weakly
• Expect to expend time, resources, and effort
• Don’t assume the first interpretation is best or complete
• Limit assessment by the amount of uncertainty, resources, and time available
• Do less assessment when you have a significant competitive advantage
• Do more assessment when failure would be catastrophic
• Expect to identify and critically evaluate the stories, narratives, and categories used to conceptualize the situation
• Do not fully trust filtered or aggregated information

Gather information

• Observe system behavior in person
• Study existing documentation, theory, and data
• Recruit and confer with experts
• Study similar and competing systems
• Define assumptions explicitly
• Identify what is missing, uncertain, or unknown
• Review responses to shocks or rapid changes
• Evaluate the accuracy and precision of data
• Free the data; identify distortion, delay, or sequestering
• Identify where power, accountability, benefits, and risks lie
• Identify reinforcing and balancing loops
• Make pictures, flow charts, and maps
• Be active in adjacent areas and projects

Encourage diversity of perspectives

• Engage a broad range of stakeholders
• Create a safe environment with standards for behavior
• Resist using power to shortcut deliberation
• Make deliberate, yet tentative boundary decisions
• Use models to encourage reflection and debate
• Share information freely and make it easy to visualize
• Seek out and maintain independent sources of information
• Engage individuals with local knowledge
• Engage outsiders with no stake in the system
• Solicit the opinions of disadvantaged or less assertive stakeholders
• Think through the problem forward and backward
• Discuss what would need to be true to support assumptions
• Prioritize utility of information over searching for objective truth
• Prioritize input from those who do the work

Assess on multiple levels

• Cycle between deep dives and big picture analysis
• Look for patterns of behavior over time
• Look for events, results, and transactions
• Look for organizing rules, priorities, and structures
• Perform both synthesis and analysis

Define purposes and values

• Define purposes and values before defining problems
• Use purposes and values to guide problem definitions
• Use purposes and values to guide boundary definitions
• State problems in solution independent terms
• Try stating the problem in both general and specific terms
• Move to a larger purpose to widen the range of solutions
• Move to a more specific purpose to narrow the range of solutions
• Make boundary decisions to construct a solvable problem
• Create narratives, categories, and stories to help conceptualize the situation