What is behavioral systems mapping?
Behavioral systems mapping (BSM) is a qualitative method that brings systems thinking and behavioral science together. A map shows who acts, what they do, and what makes those actions more or less likely. It also shows how one person’s action might affect what others do next. The aim is to explain how those behaviors combine to produce a problem and identify where an intervention could make a difference.
The approach grew out of work on the Behaviour Change Wheel (Michie et al., 2011). In 2020, Public Health England included mapping in its guidance on selecting target behaviors (West et al., 2020). Hale, Jofeh, and Chadwick (2022) used participatory mapping to study home energy retrofits in Wales. Davan Wetton and colleagues (2025) built a map from interview and focus group transcripts on waste management in Kisumu, Kenya.
Those studies illustrate how the method varies with the question and the evidence available. The Centre for Behaviour Change at University College London (2026) teaches a set of principles and methods for modeling possible causal influences in behavioral systems. A map built with workshop participants and a map built from coded transcripts can serve the same purpose while drawing on different evidence.
| Element | What it represents | Workplace example |
|---|---|---|
| Actor | A person, team, function, organization, or sector that performs an action relevant to the issue. | Executive team, project management office, department head, or project lead. |
| Behavior | An action defined precisely enough to observe directly or through a record of what happened. | Flag a dependency before confirming a delivery date. |
| Influence | A condition that changes how likely an action is, how often it occurs, or how it is carried out. | Time available, access to project data, team norms, or the response to an early warning. |
| Causal link | A hypothesis about how a change in one element contributes to a change in another. | Criticism after an early warning may lead project leads to wait longer before raising the next risk. |
| Feedback loop | A chain of causal links that feeds back into its starting point and amplifies or limits a change. | Tighter oversight leads to later warnings and more delays, which prompt tighter oversight again. |
How does BSM differ from process or stakeholder mapping?
A process map shows how activities, responsibilities, and handoffs fit together. A stakeholder map shows who has a stake in the outcome and how those interests relate. A causal systems map can connect changes in resources, demand, or other variables over time. BSM brings specific actions into that causal picture by asking who does what and how it changes the situation for someone else.
Labels such as “risk culture,” “governance,” or “poor collaboration” cover too much to serve as behaviors. Davan Wetton et al. (2025) define a behavior as an action that can be observed directly or indirectly. An influence changes how likely the action is, how often it happens, or how it is carried out.
| Map type | What it shows | Main question | Decision it supports |
|---|---|---|---|
| Process map | Activities, responsibilities, handoffs, and flows. | How is the work supposed to move from one step to the next? | Simplify a process, assign responsibilities, or repair a broken handoff. |
| Stakeholder map | Stakeholders, their interests and resources, and how they relate to one another. | Who is affected, who has influence, and who depends on whom? | Decide who to involve and how to structure decision making. |
| Causal systems map | Variables, causal links, delays, and feedback loops. | What dynamics keep the problem going over time? | Explore scenarios and identify possible leverage points. |
| Behavioral systems map | Actors, actions, influences on behavior, and links among actions. | Who does what, what shapes their actions, and what follows for others? | Choose which behaviors to investigate and potentially change. |
The maps can inform one another. A process map might locate a handoff, and a stakeholder map might identify who owns it. A behavioral systems map asks what each person actually does at that handoff and which consequences make the pattern persist. A dense diagram earns its place only if it helps someone make a better decision.
Why do organizations target the wrong behavior?
A transformation often starts with a visible gap. Managers raise risks late, teams rarely use a new tool, or people work around a safety procedure. The sponsor can see who is closest to the problem and may commission training or communication aimed at that group.
A project lead may know how to raise a risk and still wait. If reviews prize being on schedule and an early warning leads to demands for justification, waiting becomes an understandable response. A department head who filters bad news further shapes what reaches the executive team. The Public Health England guide asks designers to look for those other actors and examine whether their actions support or undermine the outcome (West et al., 2020).
How can a map shift the point of intervention?
Imagine a leadership team that wants project leads to flag dependencies before they cause delays. Risk management training is the first proposal. Looking at how project reviews work might reveal a different starting point.
The executive team asks whether the delivery date still holds. The project management office (PMO) rates projects green, yellow, or red, and a red rating triggers another review. Department heads ask project leads to resolve issues locally before raising them. Project leads wait until they have firm evidence, leaving decision makers less time to act. A missed deadline then prompts tighter oversight, giving project leads more reason to be cautious the next time.
| Actor | Observable behavior | Possible influence | Effect on the system |
|---|---|---|---|
| Executive team | Asks whether the delivery date still holds and leaves uncertainty out of the discussion. | Schedule pressure and a preference for firm commitments. | Early indications of a problem stay out of the review. |
| Project management office | Rates the project green, yellow, or red based on how far it has drifted from the plan. | Reporting rules and the need to summarize project status quickly. | An emerging risk is hard to show before it affects the plan. |
| Department head | Asks the project team to resolve the issue locally before raising it. | Concern about losing credibility or prompting more scrutiny. | The warning reaches decision makers after some options have closed. |
| Project lead | Waits for firm evidence before flagging the dependency. | Expects to justify the warning and lacks a clear mandate to raise uncertain risks. | Decision makers have less time to act, and a delay becomes more likely. |
| Senior leadership | Tightens oversight after a missed deadline. | Wants firmer commitments on future delivery dates. | Project leads become more cautious, and the loop starts again. |
The map suggests several places to act. The executive team could ask what might change the delivery date. The PMO could give teams a way to record unresolved dependencies before a milestone slips. A department head’s response to the first uncertain warning also matters because it shapes whether project leads speak up next time. Training still has a place if the inquiry shows that leads struggle to recognize which risks to escalate.
The right starting point depends on how that organization works. In another company, the bottleneck might be unreliable data or ambiguity over who owns the decision. The people who do the work can show where the proposed map fits their experience and where it needs to change.
How do you build a behavioral systems map at work?
Hale and colleagues (2022) built maps in workshops with people involved in home energy retrofits in Wales. In Kisumu, Davan Wetton et al. (2025) coded transcripts for actors, actions, influences, and links, then revised their map through two rounds of expert review. Workshops, interviews, observations, and operational records each reveal different parts of the system. The six steps below are UNREST’s proposed adaptation of those methods for organizational change.
| Step | Working question | What it produces |
|---|---|---|
| Define the issue | What outcome are we trying to explain, over what period, and within what boundaries? | A provisional boundary and a question that leaves room for more than one explanation. |
| Gather evidence | What do observations, records, interviews, and operational data tell us about how the system currently works? | Evidence with identifiable sources and several perspectives on the work. |
| Identify actors and behaviors | Who does what, in which situation, and with what observable consequences? | Specific actions that the team can identify in the work. |
| Connect the elements | What is expected to change another element, in which direction, and through what mechanism? | An initial causal map that records the basis for each link. |
| Challenge and revise the map | Which actors, links, or effects are missing? Where do accounts differ? | A revised map that records disagreements and the hypotheses to test. |
| Choose where to intervene | Which behavior could plausibly affect the outcome, is feasible to change, and has effects we can measure? | A priority behavior to assess with COM-B and an entry point for a test. |
Start with a concrete outcome and a provisional boundary. If the brief is “why project leads fail to report risks,” it may exclude the PMO and executive team before the inquiry starts. If it covers the entire company, the map can become too dense to guide a decision. In the project example, the relevant boundary extends from the first sign of a dependency to the point at which someone can decide what to do about it.
Each arrow represents a causal hypothesis. “Review meetings influence escalation” leaves the mechanism unspecified. “When a red status triggers extra scrutiny, project leads wait longer to flag uncertain risks” gives the team a direction to test. If interviews describe extra scrutiny and meeting records show some risks being welcomed early, look at when each pattern occurs and what might explain the difference.
How do you spot a feedback loop?
A feedback loop occurs when a chain of changes feeds back into the variable that set it in motion. A reinforcing loop amplifies a change. Tighter oversight can lead to later alerts, more delays, and tighter oversight again. A balancing loop limits the change. For example, a rise in incidents might trigger additional staffing that brings the incident rate back down.
In Kisumu, Davan Wetton and colleagues (2025) mapped possible links among underfunding, waste-handling practices, and conflicting views about who should manage waste. Their COM-B analysis identified physical opportunity and motivation as prominent influences in the data. Limited access to waste receptacles and affordable collection were among the practical barriers to changing how households handled waste.
A loop puts several claims in one place. Some links may draw on observations or operational records, while others come from participants’ accounts of what happens and why. Its usefulness comes from being specific enough to test. If additional review is thought to delay warnings, for example, compare how early leads raise risks under different review arrangements.
How do you choose a leverage point?
A behavior with many connections deserves attention. It may also sit outside the sponsor’s authority, be hard to change, or create problems elsewhere. A smaller change, such as how a department head responds to a first warning, may alter what several teams expect will happen if they raise a risk.
West et al. (2020) recommend applying APEASE when choosing target behaviors. The criteria cover acceptability, practicability, effectiveness, affordability, side effects, and equity. The map helps identify where the behavior sits in a feedback loop, who is affected downstream, and what a pilot could measure. A good choice accounts for what the people involved can do with the authority and resources they have.
| Question | Why it matters | What to examine |
|---|---|---|
| Would changing this behavior affect what happens downstream? | An action that influences several other actors or a feedback loop may have a wider effect. | Feedback loops, dependencies, and who is affected. |
| Does the actor have the resources and authority to act differently? | The proposed change may depend on a decision or resources the actor does not control. | Resources, decision rights, and external constraints. |
| Will the intervention be acceptable and equitable? | An improvement on average can still concentrate costs or risks on one group. | Who bears the costs, whose input is heard, and relevant APEASE criteria. |
| What indirect effects could follow? | A change may move the problem elsewhere or make another team’s work harder. | Related feedback loops, workarounds, and extra work. |
| Can we observe the effect before expanding the intervention? | A pilot can test whether the proposed link holds in practice. | A behavioral measure, the intended outcome, and a decision threshold. |
Participation helps surface links that researchers or executives might miss. It also reflects who feels able to speak. If executives are in the room, project leads may hesitate to describe what happens after they raise bad news. Give people a way to add or contest those links so the map can capture consequences that shape the work.
How do BSM, COM-B, and the Behaviour Change Wheel fit together?
Mapping helps decide which behavior to investigate. The COM-B model then helps explain why that behavior is or is not happening by examining capability, opportunity, and motivation. The Behaviour Change Wheel (in French) connects that diagnosis to possible intervention functions.
Starting with the system can change whose behavior gets assessed. A COM-B assessment of project leads, refined if needed with the TDF, may correctly identify why they delay escalation. Mapping might also identify a department head’s response to early warnings as a behavior worth assessing in its own right. The team can then focus its diagnostic work on the actions that offer a credible opportunity to change the outcome.
In the Welsh study, participants created two maps of the actors, behaviors, and influences involved in retrofitting homes. Hale et al. (2022) then used COM-B to examine which conditions the resulting recommendations addressed. The map gave the recommendations their context, and COM-B helped explain how those recommendations might influence behavior.
What tools can you use to make a map?
A whiteboard works for a first pass. Kumu can make a dense map easier to navigate, Vensim can show causal relationships and their polarity, and Loopy can illustrate a simple feedback loop. The UCL Centre for Behaviour Change includes freely available software in its BSM training (2026).
The format matters less than the record behind the map. In Kisumu, Davan Wetton et al. (2025) documented revisions after each review and linked selected variables to supporting excerpts. Keeping the evidence, disputed links, and reasons for revisions allows someone reviewing the map later to see how the team reached its conclusions.
What are the limits of behavioral systems mapping?
A map reflects a particular question, group of participants, and point in time. Changes in context can make some links obsolete, as the Kisumu researchers note (Davan Wetton et al., 2025). A map built in one division needs a fresh look before it is used in another, especially where teams have different tools or decision rights.
Detailed transcript coding takes time. Workshops can produce a working map faster, though differences in status and what participants remember affect what gets recorded. Comparing the workshop account with observed work and activity records can show, for instance, that a workaround described as occasional is part of the daily process.
A map helps prioritize hypotheses about where an intervention might work. Testing the intervention shows whether the target behavior changes and whether the intended effect travels through the system. A pilot also gives the team grounds to revise the map when people respond in ways it failed to anticipate.
What role does BSM play in organizational development?
Behavioral systems mapping keeps observable actions in view while tracing how rules, decisions, and responses make them more or less likely. That connects behavioral science at work (in French) with the organizational development work of understanding how jobs and decisions fit together. A sponsor may see, for example, that a reporting metric rewards the very delay they are trying to reduce.
In UNREST’s organizational development consulting, mapping can help choose the intervention before committing to training or changing a metric. The sponsor may discover that one of their own practices should be the first to change. What they ask for in a review, or how they respond when the answer puts the plan in doubt, can affect what the team tells them at the next meeting.
Frequently asked questions about behavioral systems mapping
What is a behavioral system?
A behavioral system is a set of actors whose actions affect one another in relation to a specific issue. It also includes the conditions that shape those actions, from access to information to the reaction a manager expects. The boundary depends on the decision the map is meant to inform.
What is the difference between behavioral systems mapping and COM-B?
BSM represents several interdependent behaviors and helps decide which ones to study. COM-B examines the conditions that shape a chosen behavior. A map might point to both how risks are reported and how they are received, with a separate COM-B assessment for each target.
Can a behavioral systems map identify the best leverage point?
It can show which options warrant a closer look and why. The choice also depends on what is feasible, who bears the costs, and what effect the organization can observe. A limited pilot tests whether the proposed change has the expected effect.
Do all stakeholders need to attend a workshop?
You can build a first map from interviews, observation, documents, or activity records. A workshop is one way to compare accounts and talk through disputed links. Separate interviews or written feedback can give people a way to challenge the map when a joint discussion is impractical.
Is behavioral systems mapping scientifically validated?
Peer-reviewed case studies document how teams have built and used behavioral systems maps, notably in public policy, health, and environmental projects. The evidence describes how the method has been applied in those settings. Whether an intervention chosen from a map improves a particular organizational outcome is a question for evaluation.
What software should you use for behavioral systems mapping?
A whiteboard is enough to sketch an initial map. Kumu suits interactive maps, Vensim supports causal diagrams, and Loopy works for simple feedback loops. Choose based on who needs to use the map, how dense it is, and how the team will keep track of revisions.
About the author
Maxime Rabéchault is the founder and president of UNREST, a change enablement and organizational development consultancy. He draws on behavioral science and industrial-organizational psychology to diagnose the conditions for change, develop observable skills, and evaluate their effects in organizations.
References
- West, R., Michie, S., Chadwick, P., Atkins, L., & Lorencatto, F. (2020). Achieving behaviour change: A guide for national government. Public Health England. Public Health England guide
- Hale, J., Jofeh, C., & Chadwick, P. (2022). Decarbonising existing homes in Wales: A participatory behavioural systems mapping approach. UCL Open: Environment, 4, e047. https://doi.org/10.14324/111.444/ucloe.000047
- Davan Wetton, J., Santilli, M., Gitau, H., Muindi, K., Zimmermann, N., Michie, S., & Davies, M. (2025). Behavioural Systems Mapping of Solid Waste Management in Kisumu, Kenya, to Understand the Role of Behaviour in a Health and Sustainability Problem. Behavioral Sciences, 15(2), 133. https://doi.org/10.3390/bs15020133
- Michie, S., van Stralen, M. M., & West, R. (2011). The behaviour change wheel: A new method for characterising and designing behaviour change interventions. Implementation Science, 6, 42. https://doi.org/10.1186/1748-5908-6-42
- UCL Centre for Behaviour Change. (2026). Principles and Practice of Behavioural Systems Mapping. UCL training module
.webp)

