Sarah Osman
I recently spoke with practitioners working at the leading edge of social and behaviour change programme design. Five organisations spanning India, East Africa, West Africa, Latin America and the UK, each with very different operating models and funding structures. Five unrelated conversations surfaced the same tensions, and those tensions tell an interesting story about where this field is.
Here is what I took away.
The gap between SBC theory and SBC practice was raised in every interview. The specifics differ by organisation, but the structural problem is identical: most implementers cannot draw a documented line between the theories they cite in proposals and the specific techniques embedded in the content they produce.
One interviewee described it as a hierarchy problem. Behavioural science expertise sits at the top of the organisation, with PhD-level directors leading design workshops. But by the time those insights reach the programme managers and creative teams making daily content decisions, the translation has degraded. Another described a donor-mandated behavioural framework that was ninety-two pages long. The writers could not operationalise it. The programme succeeded, but she suspects this happened in spite of the framework rather than because of it. A third was candid that not every content producer in her organisation could, if pressed, name which theory is operating in a given piece of work.
This is a translation problem that I pointed out before in this blog. The field has accumulated hundreds of theories. What it lacks is a reliable mechanism for converting theoretical constructs into design decisions that practitioners can make with confidence and speed.
Every organisation I spoke with had invested in bespoke digital tools: apps, platforms, gamified behaviour change products. In every case, the assessment was the same. Nothing outperformed the platforms their audiences were already using.
One interviewee said her organisation had spent heavily on custom-built solutions under funder pressure to demonstrate a digital strategy. Behaviour change games delivered no meaningful return. The only technology she is enthusiastic about is chatbots deployed within WhatsApp, precisely because they meet audiences where they already are rather than asking them to adopt new behaviour to access the intervention.
The chatbot data also revealed something no broadcast channel ever had: the way that people engage with sensitive topics based on the time they engage. It tells how people want to use technology to engage with SBC information.
The lesson is that the digital tools that work are the ones that respect existing behaviour rather than requiring new behaviour as a precondition for access. A fifth interviewee, a digital product specialist, sharpened this further. She drew a clear line between AI for production, using it to accelerate content creation, and AI for scripting, letting it generate the behavioural content itself. The first is defensible. The second is not, because the machine cannot understand humans the way humans do. She also warned that commercial AI tools are rapidly raising user expectations for chatbot quality, meaning development-sector chatbots built to a lower standard will increasingly feel inadequate.
Attribution of behaviour change to communication interventions remains, by frank admission of every person I interviewed, unresolved. The field can measure reach and recall. With effort, it can measure shifts in knowledge and attitudes. But confidently linking a communication intervention to a change in what people do remains beyond current practice.
The interviewees closest to credible approaches described combinations of longitudinal panels, social listening, and iterative content testing. These methods work. They are also significantly more expensive than standard monitoring budgets allow. Funders want evidence of behaviour change but are not willing to pay for the measurement infrastructure that would make such evidence credible.
One interviewee said the field still treats communication as a dose and behaviour change as a response. But communication operates through social systems. Until measurement frameworks catch up with the complexity of how influence works, the field will continue to struggle with measurement.
This was the most uncomfortable convergence. In multiple interviews, practitioners described a pattern in which funders had already decided on the digital tool or platform before the formative research had begun. AI and chatbot investments were particularly susceptible: the technology decision preceded the audience insight.
One interviewee described the downstream effect: co-creation processes are being hollowed out because funders increasingly pay for tools, not for the design process that makes tools appropriate. The quality degradation is invisible in short-term output metrics but shows up in whether the content resonates.
This is not a problem unique to SBC. But it is particularly damaging in a field that claims participatory design as a core commitment. If the platform is chosen before the audience is consulted, participation is performative.
I spoke to some of the most innovative actors in the field. If they are collectively struggling with the same problems, those problems are structural. They require honest institutional responses: investment in translation between theory and practice, realistic measurement budgets, and the willingness to let audience insight drive intervention design.
—
Want to find out how skilled you are at designing for change? Check this out! https://www.osmanadvisoryservices.com/skills/