Behavioral systems mapping
Also known as: Behavioural systems mapping
Diagramming the web of actors, behaviors, and influences around a problem to find high-leverage points for change.
What it means
Behavioral systems mapping visualizes a problem as an interconnected system — the many actors, their behaviors, the relationships and feedback loops among them, and the structural factors that shape them — rather than as a single target behavior in isolation. By laying out who does what, why, and how those actions reinforce or counteract each other, it reveals leverage points, bottlenecks, and unintended consequences that a narrow individual focus would miss. It draws on systems thinking and complements behavioral diagnosis by widening the frame from one person's COM-B to the whole ecology. Its challenge is complexity: maps can sprawl, and identifying which links are causal versus merely associated is hard. It matters because many stubborn problems (obesity, emissions, antibiotic overuse) are system-level, and intervening on one behavior without seeing the map often just shifts the problem elsewhere.
How a map gets built
A behavioral systems map is usually built with the people inside the system, not for them. Facilitators run workshops where participants name the actors, write each actor's behaviors on sticky notes, then draw arrows for the influences that push those behaviors up or down. Analysts sometimes code interview transcripts to seed the map instead, then bring it back to stakeholders to check. The distinctive move, relative to a generic causal-loop diagram, is keeping actors, behaviors, and influences as separate, labeled element types, so you can see who does what and what moves them. Teams render the result in free-tier or low-cost tools such as Kumu or the system-dynamics package Vensim, which let you filter a large map, trace one actor's chains, and share an interactive version rather than a static picture.
Where it came from
The named method took shape around the turn of the 2020s at UCL's Centre for Behaviour Change, which grafted participatory systems mapping onto the Behaviour Change Wheel and its COM-B model. Systems mapping supplied the habit of drawing feedback and interdependence; behavioral science supplied a disciplined vocabulary for what actually drives an action: capability, opportunity, and motivation. The synthesis matters because ordinary system maps tend to blur people, structures, and events into undifferentiated 'factors,' losing the behavioral detail an intervention needs. Pete Barbrook-Johnson and Alexandra Penn's work on participatory systems mapping set out much of the underlying craft, and treats a map as a shared model to argue over rather than a finished truth. Applied uses remain young; several published maps describe themselves as among the first of their kind.
Reading a map for leverage
A finished map is a diagnostic instrument, not decoration. You look for feedback loops that lock a bad pattern in place, for a handful of influences that many behaviors depend on, and for actors sitting at the crossing points of the most arrows; these are candidate leverage points where one change can propagate. It also surfaces the difference between intervening on individuals and intervening on the structures around them, the distinction captured by i-frame versus s-frame thinking: a map often shows the individual behavior everyone blames being held in place by incentives and constraints upstream. And it warns you off dead ends, exposing the loops that will quietly cancel a well-meant nudge or shunt the problem somewhere else in the system rather than removing it.
Limits and where it breaks down
The honest weaknesses are practical and epistemic. Maps sprawl: a rich one can hold hundreds of links, and past a certain size it stops being legible or usable. Deriving a map from coded transcripts is slow, and when stakeholders cannot help build it you lose the validation that is half the point. Above all, an arrow is a claim about causation that the exercise rarely tests; most links rest on expert judgment or association, not measured effect, so a confident-looking map can encode confident-looking errors. Maps are also stubbornly local: one built for waste in one city, or retrofit in one country, seldom transfers without rebuilding. Treat the output as a structured hypothesis to argue with and test, not as proof that a particular leverage point will work.
Examples
Mapping antibiotic overuse exposes prescribing doctors, anxious patients, diagnostic delays, and incentives together — showing that a single 'educate patients' nudge ignores the doctor and system loops.
Mapping why hospital beds stay blocked shows discharge depends on pharmacy timing, transport, social-care availability, and consultant rounds — so 'tell doctors to discharge earlier' just moves the queue.
A map of school food takes in caterers' margins, packed lunches, the length of break, and the chip shop across the road, showing why a healthy-eating assembly changes little.
Mapping why households delay loft and wall insulation surfaces landlords, installers' waiting lists, disruption fears, and upfront cost together, showing why a subsidy alone leaves the decision stuck.
A map of a city's uncollected rubbish links household sorting, informal waste-pickers, truck routes, and dumping fines, revealing that a 'please recycle' campaign ignores the collection gaps driving the behavior.
First described in Systems thinking applied to behavioral science (2010s).
Key references
- 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. doi.org/10.3390/bs15020133
- Lunetto, M., Hale, J., & Michie, S. (2022). Achieving effective climate action in cities by understanding behavioral systems. One Earth, 5(7), 745-748. doi.org/10.1016/j.oneear.2022.06.009
- Hale, J., Jofeh, C., & Chadwick, P. (2022). Decarbonising existing homes in Wales: a participatory behavioural systems mapping approach. UCL Open Environment, 4, e047. doi.org/10.14324/111.444/ucloe.000047
- Barbrook-Johnson, P., & Penn, A. S. (2022). Systems Mapping: How to build and use causal models of systems. Palgrave Macmillan. doi.org/10.1007/978-3-031-01919-7
- 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. doi.org/10.1186/1748-5908-6-42