Team meeting
Staff meetings, project syncs, decision meetings, planning sessions
Use this when three or more people from your organization are working shared items together: deciding, planning, syncing, solving. Not when it is two people (1:1), anyone outside your organization is in the room (external, even if your colleagues outnumber them), or one person delivers prepared material and the rest mostly listen (presentation; an all-staff update segment belongs there, not here). The boundary with presentation is who produces the content: if the room produces it, score here.
Roles. Led = you called the meeting or ran it. Contributed = you attended.
Weight note. Participation balance and psychological safety carries 25% and preparation and framing 15%, because dysfunctional talk (complaining, incivility) damages meeting outcomes more than functional talk helps (Kauffeld and Lehmann-Willenbrock 2012), psychological safety was the strongest of Google’s five team dynamics, and agenda presence on its own is weakly related to effectiveness (r = .14 to .18, CIPD 2023).
Debrief scores each dimension 1 to 5 against these anchors and quotes the transcript line that earned each score.
Jump to a dimension
1. Preparation and framing
15% of the scoreMeasures whether the first minutes established why this meeting, what the room will leave with, and who decides. Not whether an agenda document existed.
- 5
In the first minutes you state the purpose, the outcome (“we leave with a send date and an owner”), and who makes the call on each open item. Items are framed as questions or decisions (“do we send the letter Friday?”), not topics (“the letter”). The attendee list matches the outcome. Any pre-read went out in time to be read.
- 4
Purpose and outcome stated. One item is a topic rather than a question, or who decides goes unsaid.
- 3
A written agenda exists and is followed, but nobody says what the meeting is for or what done looks like.
- 2
Opens with “so, what’s everyone got?” or the first item on someone’s mind. Purpose emerges, if at all, midway.
- 1
No purpose, no agenda, wrong people in the room, or the first ten minutes go to logistics.
having read the pre-read, arriving with a position on the items that touch your work, and asking “what are we deciding today?” if the leader did not say.
Evidence: CIPD 2023 (goal clarity r = .51 to .54); Leach et al. 2009 (agenda works through attendee involvement); Odermatt, König and Kleinmann 2015; Hastings and Meyer 2020 (lead with context) for the language only. All sources
2. Clarity and concision
20% of the scoreMeasures whether the meeting stays on the item in front of it and whether you use short procedural moves rather than long explanations.
- 5
You name the question before discussion starts, park tangents by name (“that’s real, parking it for Thursday”), and summarize each item in a sentence before moving on. Your turns are short. You repeat a colleague’s point only to paraphrase and confirm it.
- 4
Mostly on the item. One tangent runs a few minutes before you pull it back, or one summary is skipped.
- 3
Several drifts. You re-explain points or restate others’ comments at length; the group loses the thread once or twice.
- 2
You monologue. Procedural moves are absent; the meeting is a discussion tour.
- 1
Nobody could say from the transcript which item was under discussion at a given minute.
on-topic turns under a minute, headline first, plus at least one procedural move when the leader does not make it (a time check, “we’re off the item,” a one-line summary).
Evidence: Lehmann-Willenbrock, Allen and Kauffeld 2013 (procedural statements trigger who/what/when and suppress dysfunction); CIPD 2023 (focused communication r = .28 to .59); Kauffeld and Lehmann-Willenbrock 2012. All sources
3. Close and follow-through
20% of the scoreMeasures whether every item closes as decided, assigned, or deliberately parked; whether owners read back their commitments; and whether one named person owns each open decision.
- 5
Every item ends decided, assigned, or parked with a date. In the last minutes you read every action with owner, date, and first step, and each owner confirms in their own words. Each open decision has one named decider and a decision date. You end on time.
- 4
Owners and dates on every action, but no readback, or one item is left as “we should look into that.”
- 3
Actions are scattered through the meeting and never consolidated; several lack an owner or a date.
- 2
Actions are implied, not assigned. The meeting ends when time runs out.
- 1
Soft close (“good discussion, let’s pick this up”). Nobody could reconstruct the commitments from the transcript.
confirming your own items back with a date and asking what else is needed from you before the meeting ends.
Evidence: Gollwitzer and Sheeran 2006 (d = .65); Härgestam et al. 2013 (closed-loop readback); CIPD 2023 (summarizing decisions at the end predicts effectiveness ratings); Hastings and Meyer 2020 (informed captain) for the language only. All sources
4. Participation balance and psychological safety
25% of the scoreMeasures whether talk is spread across the room, whether you sought disagreement before deciding, and whether complaining or incivility was allowed to run.
- 5
Turns are spread; no one person, including you, takes much more than a third of the talk. Before each decision you ask for the case against, by name (“Sam, what’s the argument for not doing this?”), and you thank the dissent when it comes instead of rebutting it on the spot. Quiet members are invited by name. Any complaint cycle (two or more consecutive complaint turns with no proposal) is broken within a turn (“what would we do about it?”). Interruptions, side conversations, and dismissive remarks are absent or named when they happen.
- 4
Turns mostly spread. You ask “any objections?” once, to the room in general. One quiet member never speaks. One short complaint cycle runs before you redirect it.
- 3
Two or three voices carry the meeting. You decide without asking for dissent. A complaint cycle runs several turns. One interruption or dismissive remark passes without comment.
- 2
You or one other person dominate. Dissent is rebutted or talked over. Complaining takes a real share of the meeting.
- 1
Incivility (talking over, mocking, side chat) or a sustained complaint cycle runs unchecked; people stop contributing.
yielding when you have spoken a lot, inviting a quieter colleague in, voicing the disagreement you hold before the decision rather than after, breaking a complaint cycle with a proposal, and neither interrupting nor side-talking.
Evidence: Woolley et al. 2010 (equal turn-taking predicts group performance); Edmondson 1999; Google re:Work Project Aristotle; Kauffeld and Lehmann-Willenbrock 2012 (complaining hurts more than solutions help); Odermatt et al. 2018 (incivility lowers everyone’s rating); Hastings and Meyer 2020 (farming for dissent) for the language only. All sources
5. Decision and solution focus
20% of the scoreMeasures whether items brought as decisions got decided and whether the room spent its time on proposals rather than on describing problems.
- 5
Every item brought as a decision ends decided, or deferred with a named decider and date. Proposal statements outnumber problem statements. When a problem surfaces you ask for options and criteria before debate starts. You name each decision as it happens (“so we’re going with B”).
- 4
All but one decision lands. One item cycles on the problem for a while before someone proposes.
- 3
Half the decisions land. Much of the time goes to describing problems. Decisions that are made are never named as decisions.
- 2
Most items are discussed, not decided. The meeting produces a list of concerns.
- 1
Nothing is decided, and nobody says so.
bringing options rather than only problems, stating your recommendation, and asking “so what did we decide?” when a decision is drifting.
Evidence: Kauffeld and Lehmann-Willenbrock 2012 (solution statements vs complaint cycles); Blenko, Mankins and Rogers 2010 (decision effectiveness tracks performance across more than 1,000 companies); Grant 2012. All sources
How the total is built
| Dimension | Weight |
|---|---|
| Preparation and framing | 15% |
| Clarity and concision | 20% |
| Close and follow-through | 20% |
| Participation balance and psychological safety | 25% |
| Decision and solution focus | 20% |
| Total | 100% |
Weighted = raw × weight × 20, with the weight as a decimal (a raw 5 at 25% contributes 25 points; at 15%, 15 points).
What is never scored
- Whether the decision was right. Score whether it was made, named, and owned.
- The quality of anyone’s proposal or analysis.
- A colleague’s silence on items that did not touch their work. Score whether they were invited, not whether they spoke.
- Another attendee’s incivility as your failing, unless you led the meeting and let it run.
- Meeting size, length, or attendee list set by someone else.
- Transcript artifacts and connection drops.