Systems thinking for product managers
Modern products are complex interrelated systems where a change in one feature can have a drastic knock-on effect on other parts of the product, and where the products usage can have vast unexpected consequences outside of the product on its users, non-users, communities, culture, society and the climate. Who would have thought that a social media platform could affect elections, or that a ride-sharing app would reduce drink-driving deaths.
Product Managers need to understand and utilise systems thinking in their work to help the products succeed in achieving it’s goals without causing negative impacts in the wider world.
What is systems thinking?
What is a system?
Donella Meadows, one of the original system thinkers, defined a system as:
A set of related components that work together in a particular environment to perform whatever functions are required to achieve the system’s objective
Let’s break that down:
- Components – The parts of the system.
- Working together (Interactions) – The relationships between the components and how they interact.
- Environment – The conceptual and/or physical space that the components exist and interact within.
- Functions – The actions that each component performs.
- System’s objectives – The outcomes of all the interactions of the components.
Systems can be designed or can emerge. A product is a designed system whereas weather is an emergent system. So a weather app is a designed system interacting with an emergent system, even if, in this case, the interaction is only one way as the app can’t effect the weather.
But some systems interact with other systems in more complex ways. A product that makes it easier for people to find new jobs exists within it’s own system, but it’s also going to have an effect on the labour market system, which is going to affect the GDP of countries, which is going to affect government policy decisions. And any change to any one is these systems might affect the product system, either because of the number of people looking for work or a new regulation. These are all systems within systems interacting with each other.
What makes thinking, systems thinking?
System thinking provides a perspective to consider the second and third order effects of how systems work. This way of looking at things means going beyond simple, linear thinking, cause-and-effect results, and first order effects.
Peter Senge describes systems thinking as
“a way of thinking about, and a language for describing and understanding, the forces and interrelationships that shape the behavior of systems.”
From this, we can say that systems thinking is the application of an understanding of systems, in the case of Product Managers, to the products they work with and the ecosystems they exist in.
There are lots of mental models and thinking tools and techniques that can help Product Managers understand and apply systems thinking. It’s complicated, but it’s important.
Why is it important for product managers?
Products exist in an ecosystem, not in isolation. They are always interconnected with other systems. When product managers think of their products as existing in isolation they are unlikely to appreciate the wider implications of their product decisions or consider the unintended consequences that might arise from people using the product.
Optimising for a single goal is often demonstration of linear thinking, whereas systems thinking helps the product manager be aware of the consequences of decisions, and consequences of those consequences. Life time value, a metric often used with products, doesn’t only exist from their use of the product. Many users would also consider a world without climate change to be of long term value to themselves, and so product managers should consider that when thinking about product goals.
Systems thinking solutions are often applied at a different level to the problem. For example, if a charity wanted to increase the number of volunteers it has helping (the problem), it could develop a product for recruiting volunteers (solving at the level of the problem), or it could develop a volunteering platform in partnership with other charities that sets a new standard for promoting and applying for volunteering roles so that the benefits are more impactful (solving at a different level).
How product managers can use systems thinking
Some systems thinking concepts and how product managers might use them.
Lots of product problems are bottleneck problems. Where the goal is to move users through a flow and onto a goal, there will be bottlenecks. Onboarding new users is a good example of a flow that will have bottlenecks. We sometimes call the unblocking of bottlenecks during user onboarding ‘aha’ moments, the experience a user has where the ‘get it’, they now see the value the product can provide for them. One of the most famous was Facebook’s ‘getting to 7 friends in 10 days’. Facebook realised that if a user connects with seven friends within ten days of creating an account they are far more likely to have the aha moment and so keep using the platform. Those users who didn’t connect with friends dropped off and didn’t come back. Facebook could have tried to solve the problem of not enough people joining by doing more advertising to get more , but instead product managers at Facebook optimised the onboarding journey to reduce that bottleneck.
A bottleneck is a point in a flow (of anything, from information to shipping containers) where the rate of flow is reduced so that it creates a jam behind it. Solving the problem at the point of the bottleneck has the biggest impact on flow. Reducing the incoming flow, for example, will only have a very small impact (it just means the jam grows more slowly). Increasing the flow will make the problem worse by creating an even bigger jam behind the bottleneck.
Systems thinking product managers should develop an understanding of how all the aspects of a product, user behaviour, etc., effect the goals you have set, not just on the linear path of cause and effect. But, as our system thinking teaches us, there can be unintended consequences, so it’s useful to think what else might happen if that bottleneck is removed. Does it create another bottleneck elsewhere in the system, or overwhelm some aspect and cause it to fail.
Feedback loops are a function within systems where a output from one component is used as an input for another component. The loop can be either positive, where an aspect of the system becomes amplified, or negative, where the feedback reduces a system behaviour.
A simple example might be a music player product where a user listens to jazz and then is shown more jazz songs, which they then listen to and strengthen the feedback loop. Feedback loops are used a lot in recommendation algorithms, but aren’t considered as much as complex examples where the product interacts with other systems.
Consider how Instagram might affect gym memberships and holiday travel as people want to look good for their selfies and show themselves visiting interesting places, and then how those posts create feedback loops not within the product but within the behaviour of its users. The more we see good looking people on Instagram the more we want to be good looking and post our photos, and so amplify the loop of going to the gym and interesting holiday destinations.
The iceberg model helps product managers connect the everyday events that we all experience with the underlying patterns and with the underlying systemic structure that cause the patterns and events. Of those three levels, we generally only see the events, but the patterns and structures layers are always there, below the waterline.
Examples of events are common occurrences like a queue at a shop checkout or a bug occurring in a product. We experience these as individual
Patterns are the accumulated “memories” of events, which when looked at together over time, can reveal recurring trends. For example, we might see that queues are more likely at the shop during lunchtimes because there are more customers, or that the bugs are all connected to the same feature.
Systemic structures are the lowest level of the iceberg and describe ways in which the parts of a system are organised. These structures generate the patterns and events we observe. The structures we might look for in understanding the shop queues include the working hours of the businesses, how many businesses are within easy walking distance, and the number of shop staff working at that time. And for the bugs we might look at which developers were working on that feature, and whether users are accessing another feature short;y before the bug occurs.
Not applying the iceberg model might mean a product manager sees events occurring with the product but doesn’t link what seem like unrelated events together to understand the pattern. A systems thinking product manager would not only look for the patterns but also look deeper to understand the systemic structures that are causing those patterns and events.
More systems thinking in product management
I believe that applying systems thinking to product management will increasingly become important for product managers as products become more complicated, as we realise that outcomes can be achieved by acting on systems in different ways, and as awareness of the impacts products have outside of themselves increases.
Here are some links to articles and videos about systems thinking in product management that I used in writing this post, although none of them discuss considering systems thinking in the wider context outside of the product:
- Systems Thinking in Product Management w/ Former LinkedIn PM
- Systems Thinking for Product Managers
- Systems Thinking is the Secret Weapon of an Inspiring Product Manager
- System Thinking for Product Managers by Johanna Kollmann
- Applying Systems Thinking to Your Product Development Process
- Introduction to system thinking
- Tools for Systems Thinkers: The 6 Fundamental Concepts of Systems Thinking
- The Phantom Traffic Jam – an explanation
- Chamath Palihapitiya – how we put Facebook on the path to 1 billion users