Almost nobody knows what their process is missing until it blows up halfway through. This puts it in writing before you call anyone.
A participatory process hardly ever collapses where you are looking. It collapses at the point you had not looked at, and that nobody forced you to look at.
You do not know whether your process has gaps until they blow up halfway through.
And by the time they blow up you have already spent the call, the budget and people's patience. The cost of discovering a gap grows with the calendar.
Everyone came, so the participation was a success.
The legitimacy of a process of situated collective intelligence comes from its openness, not from its representativeness. Counting heads measures something else.
You start every process from scratch, with a generic template that does not fit your territory.
Templates fill themselves in, and that is why they are useless: they produce a coherent document about a territory nobody has looked at.
A guided session, not a form. The AI walks the twelve starting points with you, beginning with the ones you see as weakest, and hands you nothing until you have actually done the work.
Territory, promoter, problem and current point. Without those no diagnosis is possible and the tool refuses to continue. It is the first gate, and it is deliberate.
One point at a time, a single question per turn. On closing each point it asks something that does not sound technical: what do you feel about the territory that you have not yet put into words. That goes into the diagnosis.
It orders what is missing by priority, points to the item that, if it falls, drags the others down with it, and for each gap formulates the next question. Never the answer.
A one-page dashboard with the twelve points, their status and, for each gap, the next question you should resolve. Not the answer: the question.
Governance: who decides if the three AMPA cannot agree, and have they said so in front of all three?
Inclusion: who cannot come at 17:00, and what second format would you run for those families?
Resources: who has the authority and the budget line to execute, and have you already spoken to that person?
The twelve points and the four states, complete. It is object A1 of the course. On the home page only the gap count is shown; here is the whole object.
If you do not give it a piece of data, it does not fill it in with something plausible: it marks it in red. A dashboard of plausible but false data is this tool failing, not succeeding.
A process that is just starting scores low, and that is fine. Mature processes do not score high: they score honest. If almost everything comes out green, be suspicious of your own answers.
The AI proposes on the basis of what you tell it; if you disagree, it adjusts. It structures and asks; it does not pronounce.
The diagnosis is worth nothing as a document. It is worth something as an agenda: if the gap is governance, the conversation is with whoever thinks they decide. This week.
Start today · €19 → See the other tools