How MeetBurn Works
Every recommendation MeetBurn makes traces back to specific, observable facts about your calendar. No black-box score — just six signals and a transparent decision process.
What MeetBurn looks at
For each recurring meeting series on your calendar, MeetBurn evaluates six binary signals — each either passes, fails, or returns “insufficient data” if there aren't enough instances to be sure.
Agenda present
Does the meeting description contain a real agenda? MeetBurn looks for numbered or bulleted lists, agenda-language keywords like “objectives” or “action items,” or a description over 50 characters after stripping out Zoom links and meeting metadata. A bare-minimum invite that only contains a video call link fails this signal. Meetings without a stated purpose are harder to evaluate, prepare for, and end on time.
Attendee count fits meeting type
Different meeting types work best at different sizes. A 1:1 should have exactly two people. A working session or design review works well with 3–8. A decision meeting tops out around 6 before consensus becomes unwieldy. An all-hands or town hall needs 9 or more to make sense as a broadcast. MeetBurn categorizes each meeting by title keywords, then checks whether the actual attendee count matches. Too many people for a working session, or too few for a broadcast, both signal a mismatch between format and function.
RSVP rate
Of the invited attendees (excluding the organizer, who is implicitly committed), what percentage accepted the most recent four instances? A passing threshold is above 70%. Low RSVP rates — where most attendees haven't responded or have declined — suggest the meeting doesn't feel urgent or necessary to the people it's supposed to involve. One caveat: this is calendar RSVP data, not actual attendance. Someone who accepted but didn't show up still counts as accepted here. The signal is best read as a commitment proxy, not a presence measure.
Series age and stability
This is a red signal — it fires when a meeting looks like it may have outlived its original purpose. Three conditions must all be true for the signal to fire: the series was created more than 180 days ago, the title has not changed across the last eight observed instances, and at least 80% of the original attendees are still on the invite. A meeting that's been running unchanged for six months with the same people is worth reconsidering — not because age alone is bad, but because healthy meetings tend to evolve as teams and goals change. If any one of the three conditions is false (the series is newer, the title changed, or attendees rotated significantly), this signal passes.
Organizer engagement
Does the person who scheduled the meeting actually attend it? MeetBurn checks the organizer's RSVP status across the most recent four instances. If the organizer accepted fewer than two of those four, this signal fails — they're skipping more than half of their own meetings, which suggests the meeting may have been scheduled on autopilot rather than out of ongoing need. If the organizer isn't in the attendee list at all (a pattern common when an executive assistant schedules on behalf of their manager), MeetBurn marks this signal as insufficient data rather than penalizing the meeting.
Decision or outcome specified
Does the meeting title or description indicate what will be produced? MeetBurn looks for decision language (“approve,” “decide,” “finalize,” “sign off”) and co-creation language (“draft,” “build,” “prototype,” “brainstorm”). A meeting called “Q3 Planning” or “Engineering Review” names a topic but not an output — it fails this signal. 1:1 meetings and all-hands broadcasts get an automatic pass, since their value doesn't depend on a single decision.
How recommendations are made
Signals 1, 2, 3, and 6 are green signals — each contributes +1 to the signal score when it passes. Signals 4 and 5 are red signals — each subtracts 1 from the score when it fails. Signals with insufficient data are scored as 0 (neutral). The range is −2 to +4.
MeetBurn then applies the following rules in order, stopping at the first match:
Score ≤ −1 AND no agenda AND no decision specified
The meeting has no stated purpose, no agenda, and clear negative signals on engagement or staleness. Something to question rather than keep.
Score between −1 and +1 AND 5+ attendees AND status/broadcast keywords in title
The meeting is a status broadcast that could be replaced with a written update. MeetBurn identifies the best async format for the specific meeting type.
Score ≥ +1 AND no other recurring meetings within ±2 hours on the same day
The meeting is otherwise healthy, but sits in isolation on the calendar — interrupting a potential focus block. The recommendation is to move it into the user's natural meeting block.
Attendee count mismatch AND agenda passes AND RSVP passes
The meeting has good engagement and a clear agenda, but the attendee count doesn't match the meeting type. Healthy meeting, wrong room size.
Duration ≥ 60 minutes AND score ≥ 0
The meeting has positive signals but runs long. A tighter timebox often preserves quality while returning focus time.
No other rule matched
Signals are mixed or healthy enough that no clear recommendation applies. Review at your own discretion.
What happens after the audit
For each meeting the audit flags, MeetBurn generates a drafted message you can edit and send. Three forms are available depending on your situation:
Question to the organiser
For meetings you attend but don't own. Raises the question without being confrontational. Framed around the specific signal that triggered the recommendation (no agenda, low attendance, format mismatch).
1:1 talking point
For raising your meeting load with your own manager. Includes cost and hours figures because this is a private conversation. Structured as a talking point doc rather than an email.
Team proposal
For meetings you own and want to change. Proposes a specific change with a two-cycle trial framing: “let's try this for two cycles, then revisit.”
The drafts use rationale framing — they explain why the meeting is worth questioning in terms of purpose, format, or attendees, not just cost. Cost figures appear in the 1:1 talking point only, where they're appropriate.
MeetBurn never sends anything. You always edit the draft and send it through your own email or Slack.
How cost and hours are computed
Annual cost is computed as: (duration + 23 min) ÷ 60 × hourly rate × attendees × instances per year. The 23-minute addition accounts for context-switching overhead — the time lost switching into and out of meeting mode. The hourly rate uses your estimated annual salary divided by 2,080 working hours, multiplied by 1.3× to account for benefits, taxes, and employment overhead (the true burdened cost to an organization). Weekly meetings use 52 instances per year, biweekly 26, monthly 12.
Annual hours are computed separately as straight meeting time: duration ÷ 60 × attendees × instances per year. No overhead, no multiplier — just real hours in the room. The cost/hours toggle in the dashboard lets you view whichever metric is more meaningful for your context. “Both” shows them side by side.
What MeetBurn doesn't do
- —Read-only access. MeetBurn only reads your calendar. It does not move, modify, cancel, or delete calendar events. Ever. The only actions it takes are generating draft text and recording that you marked a message as sent. Your calendar is always yours.
- —Never acts on your behalf. No messages are sent, no meeting changes are made, and no calendar invites are touched unless you click a specific action button. Even then, you control what happens.
- —Doesn't store meeting content. MeetBurn processes meeting titles, descriptions, attendee lists, and RSVP statuses to compute signals. It does not store meeting notes, recordings, or conversation content. Aggregate signal data is retained to power continuous monitoring; raw event data is not stored beyond what's needed for the current analysis.
- —Doesn't know what happens inside meetings. Signals are based on observable calendar metadata, not meeting outcomes. A meeting with a great agenda and 100% RSVP rate might still be a waste of time — MeetBurn can only see the outside.
Want to dig into the research?
The six signals and decision matrix are grounded in organizational research on meeting effectiveness — including work by Steven Rogelberg on meeting science, Liane Davey on team dynamics, and practitioners like Cal Newport on the cost of synchronous communication in knowledge work.