As product people, we often don’t walk in a systems wonderland. Instead, we walk through a tunnels. We’re trained to be Agile and to start small. We operate as speedboats and move away from the big oil tanker. We are trained to reduce complexity and to simplify things as much as possible.
The truth is that a lot of the products that we work on are complex and can only be simplified to a certain extent. In this post I want to touch on how systems thinking can help us solve complex problems and work on complex products.
Joe & The Juice
Think, for example, about the product manager responsible for Joe & The Juice loyalty app. Joe & The Juice is a chain of shops where people can get all kinds of juices.
As you’d expect, the product manager is laser focused on constantly improving and optimising the loyalty app. The question though is whether the PM for the Joe & The Juice Loyalty App is ignoring other aspects of the Joe & The Juice system to their detriment. Think of the customer experience in-store, the quality of the fruit or the experience levels of the ‘juicers’ in-store. The point is that all these aspects AND the loyalty app are part of the wider system that is customer loyalty.
What makes a system?
This is a great definition by Donella Meadows, one of the original system thinking pioneers. In my Joe & The Juice example, the loyalty app is one of the components in a system that sets out to achieve the objective of increasing customer loyalty. As product managers we’re used to pulling apart the bigger whole, to analyse its individual components. We need to flip our thinking when we look at systems and synthesise. We use synthesis to understand how the different components or nodes fit into a bigger whole.
In the words of Deirdre Cerminaro, Exec Director at IDEO, this focus on synthesis means that you’re “constantly jumping back and forth between the big picture and little details.” If we look at this bigger picture of a system, there are three common characteristics to highlight:
Dynamic — Because the components of systems are changing constantly, the system as a whole is changing constantly.
Evolving — Meaning that systems that have properties that will emerge from different parts of the system interacting.
Connected — As soon as you start looking into a system, you’ll start seeing how the different parts of the system are connected to each other.
If we take LinkedIn as another good example of a system and zoom out, we see that the big picture consists of several nodes. For example, the hiring that happens through LinkedIn is a system in its own right and so are the marketing solutions available through LinkedIn.
When we zoom in, we start understanding the relationships between the different nodes and the nature of these relationships, also called feedback loops.
If — for example — the number of jobs on LinkedIn increases, hiring through the platform increases (hence the plus sign).
Similarly, if my LinkedIn connections start targeting me with marketing activities, I might well reduce the number of people in my people network on LinkedIn (hence the minus).
In systems there are two types of feedback loops: reinforcing and balancing. A reinforcing feedback loop means that the cause and effect loop causes something to increase. This can be something that we want to optimise within a system, but the risk is that it will ultimately lead to instability. If the number of births per year goes up, the population will increase, which in turn will cause the number of births per year to grow.
In contrast, a balancing feedback loop is self-correcting and aims to stabilise. Instead of letting an increasing effect get out of hand, a balancing feedback loop will self-correct. We can use or create balancing feedback loops to reduce points of friction within a system.
What we’re doing in Systems Thinking is understanding the relationships between the different nodes that constitute a system. Through the Iceberg Model we can gain a below-the-surface perspective. We witness and event, and then we explore what sits underneath that; the patterns, structures and mental models that underpin the event.
If we, as an example, look at why so many social media posts feel so overproduced, we can explore the pattern of people being less authentic on Instagram and TikTok. What is structural about this behaviour? And what are the related mental models?
Applying the Iceberg Model to the “overproduced social media” event, you can then see how a product like BeReal taps into this Iceberg Model example, and has been designed to encourage people to be more authentic on social media.
The Iceberg Model is one of three ways to zoom into a system, along with feedback loops and flow diagrams. By zooming in and out of a system we can identify points of friction that we want to reduce or opportunities that we want to optimise.
Main learning point: Systems thinking enables us to understand the connections between different components, and identify opportunities to resolve points of friction within a system.
