Standing Meeting Audit

Standing Meeting Audit Skill

Standing meetings are organizational scar tissue: each was created for a reason, the reason healed, the meeting remained. Nobody audits them because each is individually defensible-ish and collectively they're the calendar. The audit tests each slot against three questions — does the original purpose still exist, does the attendance reflect it, what has it produced lately — renders verdicts (keep / shrink / merge / kill / experiment), and deploys the un-arguable tool: cancel it for two weeks and see what breaks. Almost nothing breaks, and what breaks reveals the meeting's true minimal form.

What This Skill Produces

• The inventory — every recurring meeting: origin purpose, current attendance, recent outcomes, annualized cost (via meeting-cost-meter)
• The verdicts — keep / shrink (length, cast, or frequency) / merge / kill / two-week-experiment, with reasons
• The experiment protocol — how the pause runs, what "broke" means, and the decision rule at the end
• The re-accretion guard — expiry dates on new standing meetings, and the quarterly re-audit

Required Inputs

Ask for these if not provided:

• The calendar's recurring population — the standing meetings with their casts, lengths, frequencies; the audit needs the census, not the impressions
• Per meeting, the archaeology — why it was created (ask the oldest attendee), what it decided/produced last month (check the notes — absent notes are themselves a finding)
• The dependencies — which meetings feed others (the prep meeting for the review meeting is a chain; verdicts consider chains whole)
• The authority — who can actually kill what; auditing meetings the auditor can't touch produces resentment reports, not calendars

Framework: The Audit Rules

1. Origin vs. present tense: every standing meeting was born for a purpose — the audit asks whether that purpose still exists in present tense ("the launch coordination sync" for a launch that shipped in March). Purpose-expired meetings are the easy kills, and there are always more than expected.
2. Attendance is testimony: chronic optional-izing, half the room multitasking, the same three people talking while nine watch — attendance reality testifies about value more honestly than any survey. A nine-person meeting that's actually a three-person conversation gets shrunk to the conversation.
3. Outcomes or ritual: last month's notes show decisions and actions, or they show weather. Meetings producing only status updates get the async-instead conversion test — status moves in documents; discussion needs rooms. No notes at all = the meeting itself doesn't believe it produces anything worth recording.
4. The two-week experiment beats debate: contested verdicts get the pause — cancelled two cycles, with a visible parking lot for what would have been raised. Outcome A: nothing broke → kill confirmed. Outcome B: two things broke → the meeting returns as those two things (shorter, smaller). The experiment converts opinion wars into evidence, and its reversibility is why owners accept it.
5. Guard the re-accretion: every new standing meeting gets an expiry (12 weeks default) — renewal requires the three-question test again; and the audit itself recurs quarterly at 30 minutes, because scar tissue regrows. The guard is what makes this an immune system instead of a purge.