If the storyboard helped us design better questions, the storybook asked us to design better answers. Once the research was complete, we had findings about caregivers, teachers, dairy farmers, transporters, financing, school attendance, and the surprisingly complicated mechanics of getting a packet of milk into a child’s hands consistently. A research report can hold all of that detail, and ours did, but a report existing is not the same thing as someone engaging with the evidence inside it.
So we drew again, this time creating a storybook narrated in the first person by a five-year-old named Ziwa, whose name comes from maziwa, the Swahili word for milk. Ziwa is five years old, goes to an Early Childhood Development Education center in Kenya, and for her the story of school milk is simple, some days there is milk, she drinks it, and then she gets on with the much more important business of being five.
Through her world, we tried to make the findings easier to understand while also championing the evidence. The storybook was never intended to replace the research report; it offered another door into it. Informed by a problem familiar to anyone who has tried to communicate research beyond a specialist audience, information does not automatically become communication simply because it has been made available. Behind Ziwa’s simple packet of milk sits a whole system of financing, infrastructure, coordination, and human effort, and the storybook was a way to let readers feel that complexity without drowning in it.
Recent work on scientific storytelling makes a similar case, arguing that narrative approaches can help researchers connect with lay audiences and build understanding, trust, and support for research. Communications researchers speaking at the 2025 Speaking Science conference noted that statistics tend to lose most of their persuasive grip within about a day, while stories hold onto roughly twice as much of theirs over the same stretch. Numbers convince us. Stories stay with us.
The storybook asked us to think just as carefully about how a reader would encounter a finding. Instead of starting from the default output, a report, a deck, or a brief, we started from the person who would need to use what we had learned, county officials who had to make budget decisions, program managers trying to troubleshoot supply gaps, donors weighing whether to fund scale-up, and practitioners who would never open the full report but still needed to act on its insights. For each of them, the question was not ‘What did we find?’ but ‘What are we asking this person to understand, and what are we asking them to do with it?’
Not everything needs to become a storybook; some evidence needs the space and precision of a technical report, and other audiences may need a two-page brief, a conversation, an animation, a dashboard, or something else entirely. The point is to make that a deliberate choice, not a default.
Where we nearly lost Ziwa
We introduce Ziwa as our storyteller and then, in places, lose her. She opens the book in her own voice, introducing herself, her school, and her friends, but then the chapters that follow, on program history, research methodology, county-level logistics, transporter incentives, and caregiver survey percentages, read as a fairly standard report. Ziwa appears occasionally, in captions or illustrations, and returns at the end to close the story and point readers toward the full research report. The findings are still there, the data is still there, and the evidence is rigorous, but the person we created the narrative around has become easier to lose.
Research has familiar forms, reports, decks, executive summaries, tables, charts, findings and recommendations, and those forms are useful, but once we begin working inside them, the people the research is about can recede into the background, and a five-year-old’s voice is easily drowned out.
The fix is simple to describe, even if it is hard to sustain in practice. Keep the narrator in the room for the whole book, from the cover through the closing page and everywhere between. Let Ziwa carry the county-level findings the same way she carries her own introduction, even when the data gets technical. The instinct that built the storyboard is the same instinct the storybook needed all the way through, make the person answering the question, and the person the findings are about, impossible to lose track of.
For instance, Ziwa can carry questions like ‘Why didn’t my milk arrive today?’, ‘Why is my teacher organizing its delivery?’, ‘Who is the person on the motorbike bringing it to school?’, ‘What happens when it rains and the road becomes impassable?’, and ‘Why was there enough milk before, but now there are more children and it runs out?’ These sound like the questions of a curious five-year-old, but sitting underneath them are findings about program consistency, transport infrastructure, enrollment, coordination, and the additional work teachers take on to keep school milk programs running.
Coming at it sideways
The storyboard made us think about the experience of a respondent before they answered our questions, and the storybook made us think about the experience of a reader after we had finished asking ours. Both required us to come at the work sideways. Instead of starting with the format we usually use, we started with the person. That does not mean every piece of research needs a character, a storyboard, or a narrative; it means that the format in which evidence travels is part of whether people can use it.
We would not design a research question without thinking about the person expected to answer it, and perhaps we should extend that same sentiment when we design how our findings reach the people who will use them. Evidence does not move from a report into practice on its own. They still have to reach the necessary people, hold their attention, and make sense to them. Ziwa made us realize that reaching the reader is part of the work, too.
This closes the two-part series. We started by designing questions around people and end by insisting our findings travel the same way.
