If you disassembled a juicer you could understand how it functioned by studying each individual part and how it connected to other parts. Although the juicer looks pretty complicated, it behaves in a predictable and repeatable way (unless it’s broken of course).
Compare that to this:
If we want to understand how road traffic behaves then studying the individual cars and trucks in isolation won’t help us at all.
Road traffic is an example of a complex system – one where we have to make sense of it by studying it as a whole and not by studying the individual moving parts. This is because the the behaviour of complex systems like traffic is emergent: the particular interactions of cars, drivers, weather and other factors will make the traffic behave differently from one day to the next.
Cause and effect is really interesting here. Our studies of the juicer we took apart earlier can give us a clear sense of what will happen when we press the button to activate it. We couldn’t look at a complex system like traffic and see the same predictable cause and effect.
Complexity experts, like Dave Snowden, talk about this limitation. In complex systems we can only see cause and effect after the fact. We can observe what just happened and then work out why it happened; we can’t use this insight to make a cause-and-effect rule.
We can make predictions, of course. We can’t help but make predictions; it’s in our nature. Looking at the flow of traffic over time, with sufficient amounts of observational data, we could make predictions about the likelihood of congestion or accidents but they are probabilities – they might happen – and that’s a lot less rigorous than must happen. The machinery in the juicer must do what it is designed to do – traffic is only likely to do what we predict it will do.
Complexity and your business
This discussion about traffic and juicers is directly relevant to how we run our organisations and think about work. There is a tendency in many organisations to structure and operate under the assumption that everything is repeatable, predictable and analytical like the juicer. Simply replace the picture of the juicer at the top of this post with a picture of an org chart and you’ll see what I mean:
The trouble is that organisations are complex, more like the uncertain and unpredictable flow of traffic than the predictable juicer. To make it even harder, your organisation doesn’t exist in a vacuum; it sits in an ecosystem with other organisations, customers, regulators, the environment… all interacting in unpredictable ways.
At the heart of dealing with this uncertainty is the concept of feedback loops. In particular, for your organisation, the feedback loop that happens when your organisation interacts with its customers (and competitors): you do something, customers and competitors react, you monitor and measure the new customer behaviour and sentiment, and what competitors do in response. All this data you gather should help inform your organisation on what it should do next.
The really important questions to ask yourself are how long do your customer feedback loops take to complete (from having an idea to implementing it and getting the feedback data)? How frequently do they take place? How much overhead and waste is baked into every feedback loop? How do you know you are collecting the right data to understand what’s happening?