Innovation through systems thinking

This post was written to accompany a workshop Anna Carlson and I are delivering at Service Design days in Barcelona.

Our workshop was designed to equip attendees with a general appetite for systems thinking along with a single practical skill. We chose Causal Loop Diagrams as they’re an easy to grasp but widely useful skill that can be understood fairly quickly. Also they don’t rely on study or observation.

The most commonly used examples; the left loop is balancing, the right reinforcing

Put simply, a Causal Loop Diagram (CLD) is a visual representation of any given system. They can be huge headache-inducing maps or small elegantly drawn diagrams, but in all cases they seek to represent the following:

• Nodes — things in our system
• Edges — connections between the things in our system
• Directions — in what direction the connections go
• Whether loops are reinforcing or balancing — i.e. whether they would grow (or decay) over time, or oscillate and remain balanced.

Each CLD should aim to represent a system of some kind. Often they’re used to represent a problem space with multiple moving parts. This video offers a dry but solid overview of the technique:

For the purpose of the workshop we following a simple set of steps:

• Split into managable groups of 4–6 people
• Decide on a problem that all participants will have knowledge of, something ubiquitous like the airport experience, failing to save money or insomnia.
• Start by drawing all the nodes / things / aspects of your problem space. This is not a scientific exercise so there is no wrong answer. Write each on a post-it note and explain to the group why you’ve written it and how it relates to the problem.
• Put all of the nodes on a big piece of paper and start drawing connections. What it causing what? Having them as post-its should allow you to arrange highly connected nodes more centrally and less connected nodes to the edges.
• For each connection decide on the direction (i.e. what causes what) and whether the direction is the same (i.e. as one increases the other increases) or opposite (i.e. as one increases the other decreases). If it’s the same label with a ‘+’, or if it’s different label with a ‘-’.
• Stand back and discuss as a group. Have you captured the whole space? What stands out as a particularly interesting loop? Where would you need further debate/investigation.

A CLD I created to show how reinforcing loops and balancing loops can work together

Why bother?

The process of creating a causal loop diagram is useful for a few different reasons:

• Understanding: first and foremost CLDs are useful as they provide a more three-dimensional view of whatever problem you’re looking to improve.
• Alignment: system conditions can appear subjective (until further study) so the debate about what causes what and what nodes are relevant will flush out differences in opinion in the group.
• Highlight gaps in knowledge: good systems thinking requires study of the actual system and your CLD will highlight where you need to start with observation.
• Provide starting points for change: within your CLD you’ll notice highly connected nodes and interesting causal loops. These will provide the levers for change in your system and therefore good starting points for research and design.

Further reading

This is a huge and interesting topic so I’d recommend reading some of the following:

• Systems Mapping — a big inspiration for this workshop, this takes you through a similar process to that above and introduces some other ideas.
• Using systems thinking to design better services — written by a colleague, this is a great look at combining stock-and-flow notation with service design techniques.
• Causal loop construction — a more textbook type guide to CLDs.

UPDATE: Causal loop diagrams created in the workshop

Workshop was a real success — we had 42 people most of whom hadn’t practically applied systems thinking before. They really got into their chosen problem areas and found the process rewarding. There was some difficulty in groups deciding on when to stop, and two groups were worried that their diagrams were overly negative, but I think was because they had chosen particularly gnarly areas to explore.

Here are photos of most of the diagrams — I had no shoes on as we held the workshop in the basement of the Estrella brewery and it was super hot.