Behavioural science frameworks like COM-B are not just academic scaffolding — they are the diagnostic foundation that makes the difference between a campaign that reaches people and one that changes behaviour.
Every campaign begins with a theory about how people change. It may be unstated, implicit in the choice of channel or message, but it is always there. The question is whether that theory is grounded in evidence or optimism.
COM-B — the Capability, Opportunity, Motivation-Behaviour model — is one of the most robust frameworks for making that theory explicit. Developed by Susan Michie and colleagues at University College London, it provides a structured way to diagnose why a target behaviour is or is not occurring, and therefore what needs to change for it to shift.
What COM-B actually says
The model holds that behaviour occurs when three conditions are sufficiently present:
Capability — the individual has the physical and psychological ability to perform the behaviour. This includes knowledge, skills, and cognitive capacity.
Opportunity — the physical and social environment permits the behaviour. Resources, social norms, and structural conditions either enable or constrain it.
Motivation — reflective processes (plans, intentions, evaluations) and automatic processes (habits, impulses, emotional responses) both push toward the behaviour.
The diagnostic implication is straightforward: if a campaign is failing to produce behaviour change, it is because one or more of these three conditions is insufficiently present. And crucially, the intervention you design must correspond to the actual bottleneck — not a convenient one.
Image courtesy of Alex Loup via Unsplash
The AI planning problem
When AI tools are introduced into campaign planning — for audience analysis, message generation, content production — they create significant efficiency gains. But they also introduce a specific risk: the optimisation of delivery at the expense of diagnostic rigour.
AI systems are very good at identifying patterns in data. They can tell you which messages performed best with a given audience segment, which channels produced the highest engagement, which framings reduced dropout. What they cannot do, absent a structured framework, is tell you whether any of that is closing the actual behavioural gap.
A campaign that maximises reach, engagement, and even message recall but fails to address the core capability or opportunity constraints facing its audience will produce impressive output metrics and limited behaviour change. This is not a hypothetical failure mode — it is the dominant failure mode in the sector.
Image courtesy of Alex Loup via Unsplash
"A campaign that optimises delivery without first diagnosing the behavioural bottleneck is not a better campaign — it is a faster way to reach the wrong conclusion."
How Kora integrates COM-B
Kora treats COM-B as a structural input, not an optional annotation. At the audience analysis stage (Stage 1 — Understand), the platform conducts a systematic COM-B diagnostic across target segments: what capability constraints exist, what environmental opportunities are absent, what motivational barriers are operative.
This diagnostic then directly constrains the strategy and message architecture (Stage 2 — Design). Intervention types are mapped to their corresponding COM-B components: information provision for capability deficits, social norms messaging for opportunity constraints, commitment devices or emotional appeals for motivational barriers.
At the content production stage (Stage 3 — Create), AI-generated material is evaluated against these component-specific requirements — not just for quality and tone, but for behavioural mechanism alignment.
What this means in practice
The practical consequence is that AI efficiency and behavioural rigour are not in tension. The speed gains from AI-assisted content production and channel optimisation are fully preserved. What changes is the analytic layer that sits upstream and downstream of that production — ensuring that speed is applied to the right problem.
For commissioners and programme managers, this matters because it changes the nature of the reporting conversation. Instead of reach figures and impressions, the performance question becomes: did capability increase? Did perceived social norms shift? Did motivation to act strengthen?
COM-B does not make campaign planning simple. Diagnosing behavioural constraints in a target population requires good data, experienced analysts, and honest engagement with what the evidence actually shows. But it does make the planning process honest — and that is the precondition for everything else.
Image courtesy of Alex Loup via Unsplash
More articles
See allMeasuring effects, not outputs: the challenge of IO campaign evaluation
Audience segmentation in fragile contexts: lessons from the field
From theory to practice: applying the Behaviour Change Wheel at scale
How Kora's evaluation module maps behavioural indicators to campaign objectives