# Changelog — the Millesimal Standard

Every ruling is recorded against the version of the Standard in force when the meeting
occurred. **Later versions never apply retrospectively.** That is why this file exists and
why it will never be tidied.

Changes are published with a rationale and a comment period. The Standard improves by being
argued with — see [`CONTRIBUTING.md`](CONTRIBUTING.md).

Format follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).

---

## [Unreleased] — v1.1 draft

Open for argument. Nothing below is in force, and no ruling is made against it.

### Under consideration

- **Criterion 2 (Fit) — ICP looseness.** The Standard currently says an ICP too loose to
  adjudicate against is "a finding in itself". Several practitioners have argued that needs
  to be a defined failure mode with a named outcome rather than a remark. Open.
- **Criterion 6 (Attendance) — minimum duration.** v1.0 says "long enough to hold a real
  conversation" and gives examples rather than a number. A threshold would be more
  consistently applicable and less fair. Unresolved on purpose; arguments wanted both ways.
- **Criterion 7 (Novelty) — the 90-day window.** Arbitrary, and defended only by convention.
  If a different window is right for a market, the case for it belongs here.
- **Rule R6 (held over) — the one-week interval.** Currently a judgement call. Whether it
  should be a fixed interval or a range is open.

### Not under consideration

- **Adding budget or purchase timing as criteria.** Both are claims about the future that
  nobody can verify. This is a structural position, not an omission, and it is set out in
  criterion 1. A timing condition agreed in writing before outreach began is a different
  thing and is already tested under criterion 2.
- **Graded scoring.** Binary judgments can be made consistently by different people on
  different days; graded ones cannot. Partial credit would make the Standard unadjudicable.

---

## [1.0] — 2026-09-01

First public version. In force.

### The rule

A meeting is Qualified if, and only if, it satisfies all seven criteria. There is no partial
credit.

### The seven criteria

| # | Criterion | In plain words |
|---|---|---|
| 1 | Role | Right person |
| 2 | Fit | Right company |
| 3 | Awareness | They knew who they were meeting |
| 4 | Need | They said it themselves |
| 5 | Consent | They agreed to a sales call |
| 6 | Attendance | They turned up |
| 7 | Novelty | New to the buyer |

### Rules of application

- **Objection window** — three business days, in the buyer's own timezone. Silence clears.
- **Burden** — the agency evidences the meeting; the buyer evidences any objection, against
  a named criterion.
- **R2** — the burden of an objection sits with the objector.
- **R3** — where accounts conflict, prefer the contemporaneous artefact over either party's
  later characterisation of it.
- **R6** — a genuinely borderline criterion may be held to a second sitting a week later. No
  provisional view is recorded at the first sitting.
- **Versioning** — rulings bind to the version in force at the meeting date.
- **Recordings** — no call recordings or transcripts are accepted, at any time, for any reason.

### Deliberate exclusions

Budget and purchase intent, meeting quality, and commercial outcome. A standard that tried to
measure all of these would measure none of them reliably.

---

## Versioning policy

- **Patch** (1.0.1) — wording clarified with no change to what passes or fails.
- **Minor** (1.1) — a criterion's threshold, evidence or scope changes. Published with a
  rationale and a comment period of not less than 30 days.
- **Major** (2.0) — a criterion is added or removed, or the all-or-nothing rule changes.

No version is ever applied to a meeting that occurred before it came into force. Superseded
versions stay published at their own URLs so that an old ruling can always be read against
the text it was decided under.