Doubling the attendee count doubles a meeting's real cost, even if the calendar block looks the same

What it is

A meeting's true cost is duration multiplied by the summed time cost of every attendee — double the people, double the cost, linear growth. But the calendar shows "60 minutes," not "18 people, 18 hours": the organizer sees their own hour while the team pays a day. Invitations get sent casually — adding one more attendee is a zero-social-cost safe move; leaving someone out carries the risk of giving offense. Meeting bloat derives half of its energy from this hidden price tag.

Why it happens

Three mechanisms jointly produce systematic over-invitation. First, cost invisibility: calendar and invitation interfaces display no person-hour total, and any resource lacking a price signal gets over-consumed — meetings included. Second, asymmetric social risk: the cost of inviting one extra person (diluted discussion, longer meeting) is spread across everyone and attributed to no one, while the cost of omitting someone ("why wasn't I called?") lands squarely on the organizer; the rational choice is to over-invite. Third, declining marginal value: the Nth attendee contributes less (one more perspective, one more informed party) as N grows while billing at full price — value per person-hour falls monotonically with size, which attendees experience as "an hour for the one sentence that concerned me." Added headcount also raises coordination overhead and lengthens the speaking queue, growing faster still — but even on the conservative linear accounting alone, most recurring meetings are already underwater.

Studying it

Workplace analytics over calendar and email metadata offers a ready method: total weekly meeting person-hours per team as a share of working time, correlated with self-rated productivity and fatigue — the negative association between meeting load and perceived productivity recurs across large workplace samples. On the meeting-science side, survey and diary studies examine size directly: attendee count correlates negatively with rated meeting usefulness, with ratings sagging clearly past a certain size. Intervention studies can manipulate cost visibility — displaying estimated person-hours and expense on the invitation — and compare average headcount before and after. Methodological caveats: calendar data records invitations and acceptances, not actual attention (multitasking in meetings), so it understates real time occupied and overstates its quality; and the size-usefulness correlation is confounded by meeting type (recurring meetings skew large, reviews small), requiring stratification by type before comparing.

Where it stops holding

Linear is a conservative lower bound, not an exact description: attendees' time is not equally priced (a senior member's hour costs more), and a key role's absence can void the entire meeting, with the "saved" attendance cost consumed by reconvening. There is also a class of meetings whose purpose is precisely to build shared context — an all-hands delivers "everyone heard the same thing," the multiplier itself is the product, and the linear cost is a price, not a waste. Training and relationship events are similar. The test remains the nature of the output: meetings needing everyone's voice and vote have a natural headcount ceiling; meetings needing only everyone's ears should not be run as discussion meetings.

Applying it

• Force a person-hour bill (headcount × duration, converted to expense where possible) into the meeting-invitation template, putting the price back at the point of decision.
• List attendees in two columns: required (speaking/deciding) and optional (informed), with optional attendees defaulted to minutes instead of a time slot.
• Replace all-hands recurring meetings with representation plus written recirculation: one delegate per subgroup, conclusions flowing back in writing.
• Split large gatherings in two: a broadcast briefing (recordable, asynchronous-friendly) and a small decision meeting (only those who must take a stance).
• Verification: per meeting, compute person-hours divided by decisions produced; audit any meeting exceeding twice the team median; recurring meetings failing review twice in a row get thinned or cancelled.

Related

• Same group: V5.09.1 Content that can be asynchronous should not occupy synchronous time · V5.09.2 Decisions need records, not just oral agreement
• Nearby: V1.07 Effects of Group Size · V5.11 Meeting Fatigue · V6.05 Notifications and Collaborative Noise
• Search terms: meeting cost · person-hours · workplace analytics · meeting load