Eric Berlow provides a good argument for embracing complexity and looking for simple (although I would say, specific or, better yet, measured -which is not the same as measurable) interventions. Berlow uses social network analysis to develop influence maps and identify the immediate spheres of influence of/on a policy problem.
This could potentially be used alongside the Alignment, Interest and Influence Matrix to (from a policy entrepreneur’s point of view) identify the key public or audience of a think tank’s influencing strategy.
Equally, it can help to inform the research agenda setting process by identifying researchable bottlenecks or problems that influence the particular issue under analysis.