Infant Feeding chatbot trial
Health - Design Mini Case Study
A trial for an Infant Feeding Chatbot in 2017 (Strategic Partnership)
https://abetterstartsouthend.co.uk/research-knowledge-hub/stories-of-impact/
My role: Service Design Lead
Situation
A breastfeeding support chatbot project I worked on within A Better Start Southend. The ambition was to improve access to timely, evidence-based support for parents and reduce barriers to early intervention through a digital service.
Task
I was responsible for shaping the design approach in a context where there was a clear tension between best-practice service design and the realities of clinical safety, technical feasibility, and cost. In an ideal setting, we might have taken a broader exploratory approach and iterated more extensively, but in this case the priority was to ensure design helped the team make safe, viable decisions quickly rather than becoming a lengthy parallel process.
Action
I adapted the design process to fit the level of risk and constraint rather than applying a textbook approach. I brought clinicians, engineers, researchers, and delivery colleagues together early so that feasibility, safety, and user need were considered in the same conversation, not sequentially. That meant design could act as a way of reducing ambiguity and surfacing risk early, rather than handing over concepts that proved difficult or unsafe to implement later.
I focused the work on the highest-value and highest-risk parts of the service, key user interactions, conversation flows, escalation routes, and governance requirements. I also embedded lessons from implementation and improvement science into the process, using a 'test-and-learn' approach influenced by Plan-Do-Study-Act cycles. That helped us explore assumptions quickly, review evidence as we went, and adjust the design approach in response to what we were learning from both users and clinicians.
As the concept developed, that collaborative process surfaced significant concerns around clinical risk and the cost of delivering the service to an acceptable standard. At that point, I supported pausing further design development. I saw that as an important leadership decision, ensuring design was doing its job by clarifying what was safe and viable, rather than creating momentum around a concept that the organisation could not responsibly take forward.
Result
The result was a responsible, evidence-based decision that prevented further investment in an approach that was not yet safe or viable. More importantly, design helped the team reach that conclusion earlier and with greater confidence than they otherwise might have. It enabled better decision-making, maintained trust across disciplines, and ensured design was a practical enabler of delivery and risk management rather than a bottleneck.