research
the open questions — what we don't know yet, and what would settle it
Holds the open questions that are costing you decisions, ordered by what the answer would change, and the standard for what counts as evidence. Findings land in its records.
structure generate department researchResearch — Mission
the open questions — what we don't know yet, and what would settle it.
Purpose. One paragraph. Research earns its place by changing decisions, so say which ones: the questions this department exists to close, and who is waiting on the answers. A department that produces reading nobody acts on is a subscription, not a department.
How work gets done here. Steering docs live at this root — the open questions, and what counts
as evidence. Findings land in records/; the raw material stays in memory/, which never steers.
Repeatable how-tos in processes/.
---
type: mission
standing: draft
owner: "{{owner}}"
updated: "{{date}}"
---
# Research — Mission
_the open questions — what we don't know yet, and what would settle it._
**Purpose.** _One paragraph. Research earns its place by changing decisions, so say which ones: the
questions this department exists to close, and who is waiting on the answers. A department that
produces reading nobody acts on is a subscription, not a department._
**How work gets done here.** Steering docs live at this root — the open questions, and what counts
as evidence. Findings land in `records/`; the raw material stays in `memory/`, which never steers.
Repeatable how-tos in `processes/`.
type: department
name: research
description: "the open questions — what we don't know yet, and what would settle it"
about: "Holds the open questions that are costing you decisions, ordered by what the answer would change, and the standard for what counts as evidence. Findings land in its records."
version: "0.1.0"
Also in the folder
Evidence Standarddepartments/research/evidence-standard.md
Evidence Standard
What counts as knowing something here. Written down because the alternative is that confidence is set by whoever says it most firmly.
What counts as a source. Rank them: something you measured, something a customer said in their own words, a primary document, a secondary report, a conversation someone remembers. Say where the line is under which a claim needs a caveat.
How a claim is written. Every finding carries its source and its date. A claim with neither is an opinion, and opinions are fine as long as they are labeled as ones.
Sample size, and the honest hedge. How many is enough to say "customers want" instead of "three customers said". Give the number you will hold yourself to.
What we do with a finding that contradicts canon. The most valuable thing this department can produce. Say who it goes to, and how fast — a contradiction that sits in a folder for a month has been absorbed rather than resolved.
Where the raw material lives. Notes, transcripts and exports belong in memory/, which records
and never steers. Only the distilled finding moves up.
Open Questionsdepartments/research/open-questions.md
Open Questions
The handful of things we do not know that are currently costing us decisions. Ordered by what it would change if we knew. A question nobody is waiting on is reading, not research — keep it off this list.
The questions. Numbered, each in one sentence, phrased so it can be answered rather than discussed. Under each: who is blocked on it, and what would settle it — a number, an interview, a working prototype.
How a question gets on this list. Anyone can propose one; say who agrees to add it, and what it displaces. A list that only grows is a reading list.
What we already settled. Closed questions with the one-line answer and where the working lives. This is the section that stops the same investigation being run twice.
How a finding becomes canon. When an answer stops being a finding and starts steering: who reviews it, and which document it changes. A finding that never changes a document changed nothing.