Most development programmes do their communications backwards. The field work happens for eleven months, and in month twelve someone in the head office assembles a report from whatever photos happen to be on people’s phones. The donor receives a document that proves money was spent. It rarely proves anything changed.
That is a waste, because in this sector communications is not decoration. It is how a programme survives its next funding round, recruits its partners, and keeps its beneficiaries trusting it. Here is what actually works, learned from how programmes really operate — not how brochures say they do.
Start with the donor’s rules, not your designer’s taste
Every major donor has visibility requirements: logo placement, acknowledgement lines, sometimes exact colours and clear-space rules. Ignore them and your beautiful report comes back with corrections — or worse, gets quietly shelved. Treat the visibility manual as a design constraint from day one, the way an architect treats the plot size. The programmes whose materials look effortless are the ones that built donor compliance into their templates instead of bolting it on at print time.
Design for the phone in the field, not the screen in the capital
In Pakistan, the real distribution channel for programme communication is WhatsApp. Community workers forward graphics to groups, beneficiaries save posters as phone photos, and partner staff read PDFs on mid-range Android phones over 4G that comes and goes. A poster that only works as a crisp A3 print fails the people it was made for. Test every material the way it will actually be received: compressed by WhatsApp, viewed in sunlight, read in Urdu or the regional language by someone in a hurry. If it survives that, it will survive a donor’s projector too.
Two languages means two layouts, not one translation
Urdu is read right to left, runs longer than the English it replaces, and breaks layouts designed for English first. Programmes that commission the English design and “just translate it” end up with cramped text, orphaned lines, and materials field staff quietly stop using. Budget for the second language as a design pass of its own — different line breaks, different hierarchy, sometimes a different photograph. It costs a fraction of the print run it saves from the bin.
Photograph people like people
Beneficiary photography has an ethics problem the sector keeps stepping in. Staged misery — the crying child, the grateful queue — may satisfy a reporting template, but communities recognise themselves in these images, and consent given to a visitor with a camera is not the same as consent to be a fundraising symbol. The programmes with the strongest visual libraries do three things: they get consent in the local language and in writing, they photograph people doing things rather than enduring things, and they credit and store images properly so a photo taken in year one is still findable, cleared, and usable in year three.
Make the annual report earn its reading time
Donor reports fail in a predictable way: forty pages of activities, no story of change. What reviewers actually retain is one programme, one place, one household, followed properly — what arrived, what shifted, what it cost. Structure reports around a small number of deeply documented cases, put the aggregate numbers in clean data graphics instead of paragraph soup, and write the executive summary for the person who will only read two pages. That person exists in every funding decision. Write for them first.
Build the library before you need it
The difference between programmes that communicate well and programmes that scramble is rarely talent — it is assets. A shared, organised library of cleared photographs, approved boilerplate in both languages, logo files in every format, and templates for the recurring materials (training manuals, banners, social posts, case studies) turns every new communication from a project into an afternoon. Build it once, govern it lightly, and every report, campaign and donor visit gets cheaper and faster.
Work through partners, not around them
Most programmes deliver through local partners — NGOs, community organisations, government counterparts — and those partners are your distribution network and your credibility in the room. Communications that arrive from the capital as finished, take-it-or-leave-it packages get used politely and forgotten. The programmes that communicate well treat partners as co-producers: templates partners can adapt without breaking the brand, materials reviewed by the people who will actually stand in front of the community, and feedback loops where field staff can say “this poster does not work in our district” and be believed. Practically, that means designing in editable formats, leaving deliberate space for partner logos and local detail, and testing materials in one district before printing for forty. Respect for the last mile is a design decision, not a courtesy.
Where AI genuinely helps — and where it doesn’t
Production is where AI earns its place in development communications: laying out bilingual documents, generating first drafts of routine materials, resizing one design into twelve formats, transcribing and subtitling field video. Judgement is where it doesn’t: consent, cultural accuracy, the choice of whose story leads the report. The honest model is agents for production speed, humans for every decision a community will ever see. That is the model our studio runs on, and it is why programme materials can now be produced at a pace that used to require a department.
Quick answers
What do donors actually require in programme communications?
Visibility compliance — correct logos, acknowledgement lines, disclaimers where specified — plus evidence of results. Build the compliance into your templates from day one and both get easier.
What is the most common communications failure in development programmes?
Producing everything at the end. Communications built as a year-end reporting scramble cannot capture stories, consent or photography while the work is happening. A standing asset library and monthly capture rhythm fix it.
Should programme materials be in English or local languages first?
Local language first for anything a beneficiary will read; English for donor-facing reporting. Design the local-language version as its own layout pass — translation squeezed into an English design fails the reader.
Can AI produce development-sector communications?
AI handles production well: layout, drafts of routine materials, format conversion, transcription and subtitling. Consent, cultural accuracy and story selection stay human decisions. That division is the working model, not a limitation.
If your programme’s communications are an annual scramble, the fix is not a bigger report — it is a system. Start with how we work with development organisations, or see the work itself.