A system is not the sum of its parts, but rather the product of its interactions.

First, some definitions:

• Needs: the capability required of us by our consumers.
• System: the people, processes, and tools arranged to meet a specific set of needs.
• Throughput: what the system produces to meet needs.
• Flow: the process of throughput through the system.

In systems thinking, rather than optimising for cost efficiency (using the least amount of money or time), we optimise for flow (causing the least number of blockages / wait times in the process).

A flow-efficient system can be more profitable than a cost-efficient system, due to producing more value more often and sooner, to higher quality and with lower failure demand, despite costing more along the way, and despite folks being idle throughout the process.

Analysis is the act of breaking things down into their constituent parts, and assessing them each based on their own merit.

Synthesis is the act of bringing component parts together, and assessing them based on their interactions together. We judge parts based on their contribution to that whole.

This means we might allow things that would otherwise fail an analysis, e.g. having people who are not in the constraint be idle, or to take over work from the constraint that they would otherwise not do.

Places to Intervene in a System

Donella Meadows Places to Intervene in a System. Meadows lists 9 places to intervene in a system, in ascending order of leverage/efficacy:

9. Numbers (subsidies, taxes, standards).

10. Material stocks and flows.

11. Regulating negative feedback loops.

12. Driving positive feedback loops.

13. Information flows.