Implications for Delivery
We are surrounded by complex systems such as ecological systems that support a wide variety of plant and animal life. Yet, when we design behaviour change interventions, we often focus on the impact of a single intervention without fully considering the broader system it operates in.
To deliver more effective and systemic change, consider the following:
Start by understanding systemic influences using tools like the Iceberg Model (Meadows, n.d.), which helps us consider the deeper patterns, beliefs, and mental models behind visible behaviours.
Use the Double Diamond framework (Campsall & Khatri, 2024; Design Council, 2003) to guide the process, from understanding the problem to designing and delivering a solution. This approach supports both exploration and refinement throughout the intervention design process.
Identify possible leverage points, places where a small change can have a larger impact across the system (Hallsworth, 2023). These can help you target areas that shift the system more effectively.
Finally, design with busy people in mind. Most people don’t have time to figure out complicated interventions. A useful rule of thumb is to ask:
Can you remove something to make it simpler?
Can you add something that makes it easier to understand or act on?
By thinking in systems and designing with simplicity, your interventions are more likely to be adopted, understood, and sustained over time.
Background
Achieving the goal of having the world’s safest roads by 2030 will require us to drive meaningful behaviour change across Scotland. In the road safety community, we often rely on a narrow set of behaviour change approaches. However, to create systemic change, it is essential for practitioners to draw from a range of interventions and consider how these fit into the broader systems that people live and operate in. It is essential to remember that humans make mistakes, and our behaviour change efforts must be designed with that in mind.
We already live within many systems — ecological systems (where organisms depend on one another to survive), or financial systems (which guide us through decisions like opening a bank account or managing investments). Similarly, our roads are part of a socio-technical system (Campsall & Khatri, 2024). Roads are not just physical infrastructure but shared social spaces that support everyday life and economic activity.
Instead of viewing behaviour change as a linear process where one action leads to one result, we should ask: how does this intervention fit into the wider system? For example, the widespread use of seatbelts was achieved through a system of interventions such as legislation, reminders in vehicles, public campaigns, and endorsements from influential figures.
Taking a systems approach also means applying a behavioural lens, rather than relying only on behavioural interventions to fix every problem. As the Behavioural Insights Team (BIT) Manifesto (Hallsworth, 2023) highlights: not all challenges can be solved by behavioural interventions alone, but it can be a powerful ally.
For instance, enforcement around speeding will always be necessary, but when combined with behavioural approaches such as personalising enforcement-related information, and highlighting social norms, can help in overall effectiveness.
What the Research Shows
When aiming for systemic behaviour change, it is important to look beyond the behaviour and consider the factors that come with it, such as habits. For example, Adrian Davis (2023) found that routine behaviours, like commuting, are one of the strongest predictors of driving behaviour. Behaviour is often embedded in daily routines, which are hard to shift without broader changes in the system.
In a complex system, it can be difficult to see how all the different elements interact, considering that behaviours often operate at multiple levels. For example, beliefs towards active travel may be shaped by personal habits, social influences (like peer groups), and factors enabling the behaviour such as infrastructure, time, and distance. Understanding these layers is crucial for designing effective interventions.
The Iceberg Model
One useful way to begin understanding your system is by using the Iceberg Model (Blokland & Reniers, 2020) along with other useful resources for systemic thinking from (Meadows, n.d.). The model helps explore not just what is visible on the surface (such as outcomes or behaviours), but also what lies beneath, such as the systems, patterns, beliefs, assumptions, and mental models that drive those behaviours (see Figure 1). Start from the top of the model by describing what is happening and then decode the invisible or not as visible aspects, such as the patterns of behaviours, values, beliefs and assumptions surrounding the behaviour. Meadows also introduced the bathtub analogy to explain how systems work: just like water flows in and out of a bathtub, behaviours are influenced by various inflows and outflows.
The water in the bathtub can be added to or subtracted in a variety of ways, including starting the tap for more water or opening the drain to let the water out of the bathtub. You can think of social or environmental influences as taps and drains, shaping how much of a behaviour ‘accumulates’ over time. For a deeper introduction to systems thinking and the concept of stocks and flows, see Acaroglu (2017) and Meadows (n.d.).
.png?sv=2026-02-06&spr=https&st=2026-11-17T17%3A47%3A14Z&se=2026-11-17T18%3A04%3A14Z&sr=b&sp=r&sig=ZqF%2FTe8rWZCbKDo7VV4HJS0lJ00Nl0wfttbbxAAoVXc%3D)
Figure 1. The Iceberg Model (Blockland & Reniers, 2020)
The Double Diamond Framework
Once you have an idea about the overall system in which your intervention operates, a helpful model adapted for road safety from the Design Council (2003) is the Double Diamond Framework for designing interventions (see Figure 2). The framework effectively shows a process for designing an intervention via two diamonds and four stages for arriving at a solution. It prevents us from jumping straight to a solution, without thinking through the problem and how we might address it. Often, we can decide that we need to do a campaign, for example, and then work backwards to decide what it should target. This process forces us to really understand the problem through all of the various data sources, use behavioural insights to define the problem, and then work through all of the potential elements of the intervention. It highlights when to diverge (opening up to gather ideas, or experiment with concepts) and when to converge (define the behavioural problem, and refine the intervention). It can be a useful way of thinking through a problem and developing a solution, with more detail on using the model for developing behavioural interventions available in Campsall & Khatri (2024).
.png?sv=2026-02-06&spr=https&st=2026-11-17T17%3A47%3A14Z&se=2026-11-17T18%3A04%3A14Z&sr=b&sp=r&sig=FOMnaXH8vXhavCcenlB0TEXll%2BlOXvSNptrpz%2Bix85M%3D)
Figure 2. The Double Diamond Framework from (Campsall and Khatri, 2024)
After going through this process, it is likely that you may still have several questions regarding both the problem and the desired solution. Designing with a system in mind can be complex - instead of deploying multiple interventions at once, consider if your system has any leverage points that you can consider. Hallsworth (2023) describes these as a point where a shift in behaviour will produce wider system effects. For example, your desired behaviour is to encourage more riders to wear high visibility jackets. If a small number of users start buying more of these jackets, it encourages suppliers to produce more of these, increasing supply, reducing prices and eventually increasing demand of the jackets. Thus, identifying some leverage points within your system can be beneficial for influencing the whole system, enabling wider adaptation.
Another rule of thumb to remember is that we should design interventions for busy people. Interventions should simplify decisions and behaviours, not add to people’s mental load. When trying to simplify interventions:
Can you remove something from the intervention to make it simpler and easier to engage?
Can you add something that simplifies the process or message?
While problem definition and designing a solution for a system is key, it is equally important to consider how can you evaluate a system of interventions instead of evaluating a single intervention. This broader perspective can provide more meaningful insights into what is working in the system and where to focus next.
Sources
Acaroglu, L. (2017). Tools for Systems Thinkers: Getting into Systems Dynamics...and Bathtubs. Disruptive Design. https://medium.com/disruptive-design/tools-for-systems-thinkers-getting-into-systems-dynamics-and-bathtubs-1f961f7c4073
Blokland, P., & Reniers, G. (2020). Safety Science, a Systems Thinking Perspective: From Events to Mental Models and Sustainable Safety. Sustainability, 12(12), Article 12. https://doi.org/10.3390/su12125164
Campsall, D., & Khatri, D. (2024). Behaviour in the Safe System [Whitepaper]. Agilysis Limited. https://agilysis.co.uk/publications/#173-165-wpfd-white-papers-reports
Design Council. (2003). Design Council. The Double Diamond. https://www.designcouncil.org.uk/our-resources/the-double-diamond/
Hallsworth, D. M. (2023). A Manifesto for Applying Behavioral Science [Manifesto]. Behavioural Insights Team (BIT). https://www.bi.team/publications/a-manifesto-for-applying-behavioral-science/
Meadows, D. (n.d.). The Donella Meadows Projects Academy for Systems Change. Systems Thinking Resources. Retrieved 3 June 2025, from https://donellameadows.org/systems-thinking-resources/
Prof Adrian Davis. (2023). Essential Evidence 4 Scotland No.69 Can changing social norms change travel behaviour? (No. No. 69; Essential Evidence 4 Scotland). Transport Research Institute. https://blogs.napier.ac.uk/tri/essential-evidence-scotland/?utm_source=localtransporttoday&utm_medium=newsletter&utm_campaign=essentialevidence25&utm_term=quarterpage_marchedition&utm_content=shorturl