Participatory Consultation
How it works
In conflict-affected places, including large or remote living communities in dialogue processes is difficult - the geography is fragmented; and the usual workarounds quietly narrow whose voice actually counts. Written surveys exclude anyone less literate and flatten nuance into multiple choice answers.
This consultation runs over WhatsApp and hands the reading to AI sensemaking. People answer open questions with a voice note or a typed message, in whatever language they speak. The analysis then maps the themes participants raised, counts each distinct idea across the group, and finds the cruxes where opinion splits.
01 - Ask
A consultation goes out over WhatsApp. People answer at their own pace, in their own language, via text or voice message.
02 - Gather
In practice, most responses arrive as voice notes. Each is transcribed and translated on arrival, the original kept alongside, and anonymised before anyone reads it.
03 - Cluster
The analysis begins by reading everything at once and mapping themes and sub-themes out of what was actually said. A category exists because participants put it there, and the most discussed themes come first.
04 - Extract
A single answer often carries several distinct ideas. Each becomes its own claim, tied to the exact quote it came from; similar claims are then merged and counted across all participants, so the report can say how many people made a point, and in whose words.
05 - Cruxes
Beyond what most participants agree on, the analysis tests which statements split opinion most sharply and where each participant stands on them. These cruxes are often the most useful finding: they mark the questions a dialogue actually needs to address.
06 - Report
The result is an interactive report in which every finding carries its citations: open a claim, and the quotes behind it are there. It goes to the team running the process and back to the participants, who see what the group as a whole said.
Insight in days, not monthsSee it run