Your calendar is a bad witness to your meeting problem. It shows what someone booked. It doesn't show the standup that ran 25 minutes, the "quick call" nobody scheduled, or the review you sat through with your camera off, writing code in another window.
That's why most advice about too many meetings misfires. It tells you to cancel things before you know which ones cost you the most.
Short answer: if you have too many meetings, audit one week before you cut anything. (1) List every meeting from your calendar. (2) Compare it with how long the calls actually ran. (3) Price each recurring one (salary × 1.4 ÷ 2,080 per attendee-hour). (4) Mark it keep, shorten, make async or cut. (5) Send a short script for each change. The research says both the count and the quality of meetings matter, and that gaps between calls go with less fatigue.
Signs you have too many meetings at work
How many meetings is too many? I couldn't find a study that sets a weekly limit; the caps quoted online cite none. What research does show is that both the count and the quality matter.
In a one-week diary study, Luong and Rogelberg (2005) found that more meetings went with more fatigue and a heavier sense of workload that day. In two surveys of full-time employees, Rogelberg et al. (2006) found that how much meeting time hurt depended partly on meeting quality, and that perceived meeting effectiveness had a strong, direct link to well-being. Fewer bad meetings beats fewer meetings.
Of 182 senior managers surveyed by Perlow, Hadley and Eun (2017), about two-thirds (65%) said meetings stop them finishing their own work, and 64% said deep thinking pays the price. So the practical signs are about what meetings push out:
- Your own work moves to the edges of the day. Perlow's team describes people coming in early, staying late or using weekends to find quiet time.
- Most days have no two-hour block. At one consultancy they studied, very few two- or three-hour blocks were left.
- You accept invites you'd rather decline. People in a 2022 survey wanted to decline 31% of invitations and declined 14% (Otter.ai and Rogelberg, 2022).
How much time meetings take, and which numbers to trust
It depends on who is measured and how. Microsoft's telemetry is the largest dataset: in the 2022 Work Trend Index, the average Teams user's weekly meeting time was up 252% since February 2020. The 2023 report put the heaviest quarter of meeting users at 7.5 hours a week. Self-reports run higher: 18 hours a week in the Otter.ai survey. Perlow's article puts executives at nearly 23 hours a week, without saying how that was measured.
Three cautions before you quote any of them:
- Teams numbers count Teams. A 2019 meeting in a conference room never touched Teams, so part of that jump is meetings moving onto video. That's my reading of the method, not Microsoft's claim.
- "Three times more" is a count. The 2023 report said people were in 3x more Teams meetings and calls per week; Forbes (2024) reported it as meeting time tripling.
- Samples skew. The Otter.ai survey drew 632 volunteers with a median salary of $180,000; a third were SVP or above. Microsoft's "interrupted every two minutes" is computed on the top 20% of users by ping volume, per its methodology note (2025).
The number that matters is yours. Getting it is the audit.
How to run a one-week meeting audit
Five steps over one ordinary week, not one with a launch or a holiday in it.
Step 1 — List the week. Export your calendar as an .ics file (paths checked October 2026). In Google Calendar, point to the calendar under My calendars, then More ▸ Settings and sharing ▸ Export calendar (Google Calendar Help). In Calendar on a Mac, click the calendar and choose File ▸ Export ▸ Export (Apple Support). Drop the file into the free calendar audit, which costs every meeting in the week and projects the year. Mark the recurring ones.
Step 2 — Compare it with what happened. A calendar misses overruns, cancellations and calls nobody booked, and that last group is big: unscheduled ad hoc calls were 64% of all Teams meetings in Microsoft's 2022 data. The free method is a tally: write the real start and end beside each meeting, plus a line for every unplanned call. Tedious for five days, and it works.
The way I'd do it (I build Timex, so weigh that) is to let the Mac keep the record. Timex logs the frontmost app, window title and active browser tab's URL every second, and its idle gate keeps counting while the app in front plays audio or the mic is live, so an hour of listening on a call stays an hour (how Timex counts calls).
Add a Meetings category, right-click Zoom or Teams to move them into it, add a website rule for meet.google.com, pick Last 7, and the chart stacks meeting time per day. A meeting in a room, or a phone call away from the desk, appears as an away gap of five minutes or more, which you can turn into a Meetings entry.
Timex doesn't read your calendar, and here that's the point: it's a second witness.
Step 3 — Price each recurring meeting. Hourly cost per attendee is salary × 1.4 ÷ 2,080, the formula behind the free meeting cost calculator. The 2,080 is 52 weeks of 40 hours. The 1.4 is the benefits load, and it holds up: wages and salaries were 70.0% of private-industry compensation costs in June 2026 (BLS, 2026), so total compensation runs about 1.43 times pay. The real cost of meetings is higher, because the formula ignores the work that didn't get done.
An example team where everyone earns $130,000, or $87.50 per person-hour:
| Recurring meeting | Length × people | Cadence | Per meeting | Per year |
|---|---|---|---|---|
| Standup | 15 min × 8 | daily (250 days) | $175 | $43,750 |
| Status update | 60 min × 10 | weekly | $875 | $45,500 |
| Sprint planning | 90 min × 8 | every 2 weeks | $1,050 | $27,300 |
| 1:1 with manager | 30 min × 2 | weekly | $87.50 | $4,550 |
That's $121,100 a year for four unremarkable meetings. Now apply Step 2: if the standup really runs 25 minutes, it costs $72,917 a year, and the overrun alone is $29,167.
Step 4 — Keep, shorten, make async or cut. Ask the attendees, not only the organizer: Rogelberg (2019) reports that leaders rate their own meetings much more favorably than attendees do.
| Decision | When it fits | Evidence | In the example |
|---|---|---|---|
| Keep | Needs live back-and-forth, or it's relationship time | Effectiveness tracks well-being (Rogelberg et al. 2006); satisfaction fell where meetings went to zero (Laker et al. 2022) | The 1:1 |
| Shorten | Routinely outlasts its purpose | Seated meetings ran 34% longer than standing ones, with no better decisions (Bluedorn, Turban and Love, 1999) | Planning to 60 min saves $9,100 a year |
| Make async | Information flows one way | 71% said good notes would make them more comfortable skipping non-critical meetings; they want decisions, dates and action items (Otter.ai and Rogelberg 2022) | Written status, plus 15 min on Thursday only if a decision is flagged: saves at least $34,125 a year |
| Cut, or decline | You aren't needed for the decision | People in meetings they felt unneeded in reckoned they multitasked 70% of the time (same survey) | Anyone invited just to listen |
Step 5 — Send the scripts. The same survey shows why people don't decline: 90% feel comfortable declining over a schedule conflict, only 51% when they're just too busy, and 78% say their manager hasn't explicitly talked about declining. So start with your manager, and when you decline a meeting, always give the organizer a way to stay covered.
- To your manager: "Can we agree which of my recurring meetings I can skip if I get the notes? I want two afternoons a week clear for [project]."
- To an organizer: "I don't think I'm needed for the decision on this one. Could you send me the decisions and action items after? If there's a question only I can answer, put it in the doc and I'll reply before you meet."
- To your team: "For four weeks, let's post status in [doc] by Monday 10:00 and drop the Monday call. If someone flags a decision, we take 15 minutes on Thursday. In week five we check what slipped."
The trial-then-check shape is deliberate; it's the measure-and-debrief loop Perlow's team recommends.
Why back-to-back meetings wear you down
The changeovers cost something, not just the hours. Microsoft's Human Factors Lab (2021) put EEG caps on 14 people for four 30-minute video meetings, once back to back and once with 10-minute meditation breaks, swapping the order a week later. Without breaks, beta-wave activity, which the researchers associate with stress, climbed across the two hours and spiked at each transition. With breaks it held steady.
Read it as a demonstration, not proof: 14 people, a company lab, breaks spent meditating rather than answering email, and results that fed an Outlook feature. Larger survey data points the same way: in a survey of 9,787 people, Fauville et al. (2023) found more time between video meetings went with less fatigue.
Separately, Allen et al. (2022) surveyed working adults about their last meeting and linked meeting recovery, the time it takes to get into your next task, to how satisfying and effective that meeting was. Jumping from a call into focused work adds the cost of context switching.
The fix is a default, because nobody ends meetings early by willpower (menu paths checked October 2026):
- Outlook: in new Outlook, Settings ▸ Calendar ▸ Events and invitations ▸ Shorten duration for all events; in classic Outlook, File ▸ Options ▸ Calendar ▸ Shorten appointments and meetings (Microsoft Support).
- Google Calendar: Settings ▸ General ▸ Event settings sets the default length of new events (Google Calendar Help); Google's 2020 tip points to a Speedy meetings box there.
Start late rather than end early. In that tip, a G Suite product manager concedes that "even if a meeting is scheduled to end at, say, 4:25, it tends to run until 4:30." Microsoft's example does it this way: an 11:00 half-hour becomes 25 minutes starting at 11:05.
Meeting fatigue on video calls
Video adds its own load. Bailenson (2021) proposed four causes: hours of close-up eye contact, the extra work of sending and reading cues through a screen, watching your own face, and being stuck inside the camera frame. He calls them "arguments, not yet scientific findings."
Fauville et al. (2023) surveyed 9,787 people in early 2021 with the 15-item Zoom Exhaustion & Fatigue (ZEF) scale. More frequent and longer video meetings went with more fatigue, and so did mirror anxiety, feeling stared at, feeling trapped and the effort of managing cues, even after accounting for how much people used video. Women reported more fatigue, which the authors trace to those same mechanisms. The limits: a self-selected sample, self-report, correlation only, mid-pandemic.
The cheap fixes follow from the causes: hide self-view, make calls that don't need faces audio-only, and sit back from the camera, with an external keyboard and webcam if you have them.
Did cutting meetings work when companies tried?
Partly, and only where rules stopped the calendar from refilling. In January 2023 Shopify deleted recurring meetings with three or more people, imposed a two-week cooling-off period, brought back meeting-free Wednesdays and confined meetings of over 50 people to a Thursday window (CNN, 2023).
In July 2023 the CFO told Fortune (2023) that time per person in meetings over January to May was down 14% on the same months a year earlier, and Wednesday meeting time down about 26%. In September the company told Fast Company (2023) it was on track to ship 18% more projects.
Those are company figures with no comparison group, a year after Shopify cut 10% of its staff. On 4 May 2023, inside that January-to-May window, Shopify also cut 20% of its workforce, in cuts that appeared aimed at manager roles the CEO said add "heavy layers of process, approvals, meetings" (CBC, 2023). So neither the 14% nor the 18% belongs to the meeting purge alone.
And 14% is far from an empty calendar. Judd Antin, who ran Airbnb's design studio, wrote (2023) that Airbnb's calendar amnesty refilled almost immediately. Shopify kept its rules and added a cost estimate to invites for three or more people.
Laker et al. (2022) surveyed 76 large companies with one to five no-meeting days a week and reported the best results at three. Treat its percentages as directional: the firms picked the policy themselves, outcomes come from pulse surveys, and the same article claims knowledge workers spend "more than 85%" of their time in meetings, far above every time-use figure in this post.
A purge buys a short break. A rule makes it last.
The honest limits
Much of this evidence is self-report, and a lot of it is sponsored. Otter.ai sells meeting notes, Microsoft sells Outlook and Teams, and Shopify grades its own homework. The peer-reviewed studies show associations; none shows that cutting your meetings will raise your output.
The cost formula prices attendance, not value. A $1,050 planning session that prevents a month of rework is cheap. Use the number to start a conversation, not to end one.
Timex measures you, not your team, and only on a Mac. It counts a call while the call is the app in front; take notes in another app with Zoom behind it and that time lands on the notes app, or goes idle if you're only listening. Phone calls and meeting rooms appear only if you label the away gap. It doesn't know who was on the call or whether anything got decided, and there's no team dashboard.
Run it next week
Export the calendar, keep the tally on paper or in Timex (the 100-hour trial is about two and a half working weeks), price the recurring meetings, and send one script by Friday. Even if you change nothing else, you'll know which meeting is the expensive one.
— DK