Making a research portfolio legible to two governments.

The situation

A collaborative research program spanning dozens of projects between two countries, across multiple disciplines and two national policy contexts.

The problem

A portfolio at that scale produces a familiar failure. Every individual project is documented thoroughly, and the portfolio as a whole is invisible. An officer who needed to understand what the program was achieving in a given area had no way to find out short of reading everything.

Underneath that sat a second problem. The two primary audiences did not share a frame of reference. One understood the work through its own program logic; the other through its own national development priorities. A document written for one would fail quietly with the other.

The approach

Portfolio level synthesis anchored to the host country's national development priorities, so its officials could locate the work inside their own framework rather than an imported one. Named impact pathways running through the narrative, so individual projects read as parts of a pattern rather than a list.

Consistent project formats with geography and thematic tagging, so comparison within a theme became possible and officers could extract the projects relevant to their brief without reading the rest.

Language throughout treated the host country's institutions as co-producers of knowledge rather than recipients of it, which is both accurate and materially different in how it lands.

The outcome

The portfolio became something officers could use rather than something they had to read. The plain language held throughout, avoiding the abbreviations and sector shorthand that development reporting normally depends on and that quietly excludes anyone outside the program.

Inclusion considerations were woven through the project narratives rather than confined to a section of their own, which turned that material from a reporting requirement into part of the case for impact. The approach is not specific to development: any organisation reporting to a funder it does not control, in a language it did not choose, has some version of the same problem.