Dr Grainne Dickerson discusses how to create effective campaigns using behavioural science. Dr Grainne Dickerson is Magpie’s Director of Behavioural Insights and Chair of The Behavioural Science and Public Health Network @BSPHNetwork.


Behavioural science is being used by an ever-growing number of campaigners, communicators and policymakers to understand why people and communities behave in the way they do.

It considers the cognitive, social and environmental drivers and barriers that influence behaviour and helps to build an evidence base about what influences behaviour at individual, community and population levels. This evidence base provides insights into what can be done to improve the design of policies, programmes, communications, products and services aimed at achieving better social outcomes for individual people and society as a whole. 

At Magpie, all behavioural scientists hold PhDs in this field and work on projects relating to improving health and wellbeing, tackling climate change, promoting equality, diversity and inclusion and poverty reduction. 

We’re often asked how we guide the transition from identifying a problem to developing a campaign or intervention based on behavioural science principles. In this guide, we explore 8 steps to achieve this:

Step one – Make it behavioural:

Magpie’s clients are faced with many difficult challenges often stemming from problems such as rising rates of obesity, low levels of physical activity, low compliance with COVID-19 rules, overuse of antibiotics, just to name a few. These issues are often framed as problems and our job is to work with clients to do two things:

  1. Identify target audiences: Gaining an understanding of who we need to address these problems with. This could be the people who need it most, the people who are most likely to change initially, those who are most difficult to reach or all of these groups. 
  2. Identify target behaviours: When we understand the problem we are trying to solve, we can begin listing the new behaviours that the target audience would need to do to address the problem. This list of behaviours then needs to be prioritised according to what people are mostly likely to change.


  • Translate the problem into a comprehensive description of the target audience and their behaviours.
  • Prioritise which behaviours to start changing.

Step two – Don’t reinvent the wheel:

Above: A diagram showing how the behaviour change wheel uses policy categories, intervention functions and sources of behaviour to design interventions.

Existing published literature and research can often tell us about the target audiences and their barriers to engaging in specific behaviours. Many organisations and clients will have data like this already, though it might be unstructured. We combine this insight with existing research, academic studies and literature.

Some data will be available openly such as the public health fingertips data published by the UKHSA or local Joint Strategic Needs Assessments. Other data won’t be so openly accessible and may have to be actively sought out and accessed.

All of this provides a level of intelligence to get us started in a strong and scientific way. It keeps us ethical too, we shouldn’t be asking the public to engage in research if the answers already exist. 


  • Utilise existing published literature to understand target audiences and their barriers to specific behaviours.
  • Combine insights from literature with available data, both openly accessible (e.g., public health fingertips data) and potentially obtainable through targeted questioning, to gather intelligence and ensure an ethical approach that avoids duplicative research efforts.

Step three – Be theory driven:

We know that research and intervention design is more effective when it is based on behavioural theory. As part of our research, we identify behavioural theory that applies to the audiences and behaviours identified in step one.

Above: A diagram representing the COM-B Model, which demonstrates how three interconnected components: capability, opportunity and motivation impact behaviour.

The COM-B Model is a particularly useful evidence-based model for unpacking and understanding a diverse range of human behaviours. COM-B stands for Capability – Opportunity – Motivation – Behaviour (Michie et al., 2011; 2014). The COM-B model is used to unpack behaviour in the context in which it occurs. According to the COM-B model, the following must be in place for behaviour change to occur: 

  1. Capability: People must have the capability to do the behaviour. Specifically, they must have the physical strength, knowledge, skills, stamina etc to perform the behaviour. 
  2. Opportunity: there must be the opportunity for the behaviour to occur in terms of a conducive physical and social environment. For example, the behaviour must be physically accessible, affordable, socially acceptable, and there must be sufficient time.
  3. Motivation: In order for behaviour change to occur, people must also be sufficiently motivated. Specifically, people must be more motivated to do the behaviour at the relevant time than not to do the behaviour, or to engage in a competing behaviour.

We can also use the COM-B in combination with other relevant theories. For example, we’ve recently drawn on Protection Motivation Theory to understand how people respond to threat information and Self Regulation Theory when understanding decisions about medication.


  • Apply the most suitable behavioural theory to the target audience and desired behaviour change.
  • Use theories like the COM-B Model, Protection Motivation Theory and Self-Regulation Theory to guide research and intervention design.

Step four Fill gaps in understanding:

Once we’ve gathered this information, we can see what (if any) gaps exist in our understanding. These gaps can be filled by conducting field research and devising the right questions to ask of the right people. If there are gaps, this is the time to develop research questions, which will then lead to the development of research methodology. 


  • Identify if primary research is necessary to gain a complete understanding of the problem.
  • Develop research questions and methodologies to address any gaps in knowledge.

Step five Ensure feasibility:

Research is almost always constrained by money and time. When we design research methodology, we start with the methods that give us the best opportunity of answering our research questions. Research methods are simply the methods you will use to conduct your research. These are largely qualitative, quantitative or mixed methods where both are used.


  • Devise research methodologies that fit within the constraints of time and budget.
  • Choose qualitative, quantitative, or mixed research methods as appropriate.

Step six Distil actionable insights:

Now that we have the research data (secondary data that we’ve collected and primary research), we can translate the research findings into action. The concept of a behavioural diagnosis was put forward by Michie and colleagues (2011) when they showed that a COM-B Model can diagnose what needs to change for behaviour change to occur. Magpie uses behavioural diagnosis to translate the research findings into actionable insights for campaign or intervention design.


  • Analyse research findings, including secondary and primary data, to extract meaningful insights.
  • Utilise a behavioural diagnosis, such as the COM-B Model, to translate research into actionable insights for campaign or intervention design.

Step seven – Intervention design:

The behavioural diagnosis serves as a great tool to guide a multi-disciplined team working together to design an intervention. At Magpie, our behavioural insights team works with the creative designers and marketing and communications consultants to design the intervention. Once the creative team starts creating concepts and ideas, we keep checking back and iterating our ideas. It’s always rewarding to devise a strategy for a campaign and then see it come to life. 


  • Employ a co-production approach by involving a multi-disciplinary team, including behavioural insights, creative designers and marketing consultants, to design the intervention.
  • Continuously iterate and refine ideas throughout the design process.

Step eight – Keep a behavioural science thread:

Each time you design an element of your intervention or campaign, keep linking back to your behavioural diagnosis. For every element, keep track of what it’s addressing. This will make sure your intervention is grounded in your behavioural science findings.


  • Consistently refer back to the behavioural diagnosis during intervention or campaign design.
  • Ensure that each element of the intervention aligns with the behavioural science findings.

Followed correctly, these eight steps will help you to successfully transition from behavioural insight to behaviour change. 


Get in touch with us on hello@wearemagpie.com and we’ll be happy to help!