As we recognize the interconnectedness of our worlds we look for more appropriate ways to approach our "problems" and our wishes. The iconic two step diagnose and treat model fails us too often for comfort.
When my simple handy heuristics (rules of thumb) fail, then I draw Graphs of the problem situation.
With a graph in hand I can have exploratory conversations with myself or with others, clarifying, modifying, extendeing, weighting the graph. Ideally the nodes and the connections are named. I find pen and paper are the best tools to start with. If something interesting shows up then I move to computer graphic and there are a large number of adequate applications. I would urge the use of tools that exceed the capability of the popular tree diagramming tools ("mind mapping"--which are not very much like mind maps at all). A good place to start is Arrows Tool by Neo4j 
]. After that might be Plectica . Their educational material explaining DSRP (distinctions, systems, relationships, and perspectives) is definitely worth study.
Once I have a graph to work with, I look at it through Dave Snowden's Cynefin frame work, identifying the level of complexity for subgraph (parts of the problem situation). This allows me to use the right typed of approaches for the type of problem (simple, complicated, complex, chaotic, disordered).
For socially complex problems I then use the Pierson-Meadows Nested Questions: When, Where, Why, Who, How and finally What. Consider using Six Questions Arrows Template. This is done with key stakeholders and requires two to four hours. I have added Time and Place to contextualize Meadows 12 leverage points (in four groups of three).
Once we have a handle on the **situation type** and **situation scale**, I like to use William Smith's balanced AIC conversations, spending equal time appreciating the whole of the situation, the diverse valuations in the situation based upon stakeholder positions, and finally the design-production-control-implementation conversations. The magic is in keeping the attention on all three domains so that the most Power (all three types of power) possible can be applied to the resolution-creation process.
For Complicated Problems I like to use design and engineering tools and methods. Fredrick Vester's Sensitivity Model explained in "The Art of Interconnected Thinking" is available (BrainSite ) but requires expertise and substantial time commitment.