TL;DR: They work, and there is a real study behind the claim rather than a vendor blog. Laker, Pereira, Budhwar and Malik surveyed 76 companies, each with more than 1,000 employees and operations in more than 50 countries, that had introduced between one and five meeting-free days per week over the preceding twelve months. Adoption clustered at two days, chosen by 47 percent of them, with 35 percent going to three days, 11 percent to four, and 7 percent eliminating meetings entirely. At two meeting-free days, a 40 percent cut in meetings, self-reported productivity rose 71 percent and satisfaction 52 percent. At three days, a 60 percent cut, cooperation improved 55 percent and stress risk fell 57 percent. At four days, the tendency to micromanage fell 74 percent, engagement rose 44 percent, and remaining communication was rated 65 percent more transparent. At five days, with meetings gone completely, satisfaction, productivity, engagement and cooperation all declined. The authors concluded three days is optimal and that benefits start to shrink past a 60 percent reduction. What nobody appears to have done is check that curve against how many meeting hours are actually wasted. Against an independently measured baseline of about 18 meeting hours per week, of which roughly 6 are genuinely skippable, a two-day policy already removes 7.2 hours, more than the entire waste pool. That single comparison explains the shape of the curve, tells you where your own limit sits, and is the reason most of the policy’s failures are failures of relocation, not of principle.
The question gets asked in a particular way, usually by someone who has been handed a proposal and is not sure whether it is a serious operating change or a wellbeing gesture. Does blocking out days where nobody can book a meeting actually produce anything, or does the same work just get compressed into the days that remain?
The honest answer has three parts. There is real evidence, it is more favourable than sceptics expect, and it has a ceiling that the companies publicising the policy rarely mention.
What the study actually measured
The primary evidence is a survey by Benjamin Laker, Vijay Pereira, Pawan Budhwar and Ashish Malik, published in MIT Sloan Management Review in January 2022. They surveyed 76 companies, each employing more than 1,000 people and operating in more than 50 countries, all of which had introduced weekly meeting-free days during the previous twelve months. They interviewed managers and HR directors at each, compared employee stress levels before and after, and tracked productivity, collaboration and engagement through pulse surveys.
Two things about the method should shape how much weight you give the numbers.
The outcomes are self-reported ratings from pulse surveys, not measured output. A 71 percent rise in productivity means the ratings employees gave on a productivity item rose by 71 percent. That is a real signal about experience and a weak signal about units of work produced. Anyone quoting this as “productivity increased 71 percent” without that qualification is overselling it.
The design is also a before-and-after comparison inside companies that chose to adopt the policy, not a randomised trial. Companies that introduce meeting-free days are not a random sample of companies, and the people answering the pulse surveys knew a popular change had been made.
With those caveats stated, the pattern across 76 organisations and five dosage levels is still the best evidence available, and its most useful feature is not the headline number. It is that the curve turns.
The dosage curve, and where it turns
Table 1. Reported effects by meeting-free-day dosage, with the meeting hours each dosage removes, CEOtudent editorial framework combining published survey findings with an independently measured meeting baseline
| Meeting-free days per week | Meeting reduction | Hours removed from an 18-hour weekly baseline | Share of a 50-hour week returned | Reported effects at this dosage |
|---|---|---|---|---|
| 1 | 20% | 3.6 h | 7.2% | Autonomy, communication, engagement and satisfaction improved; micromanagement and stress fell |
| 2 | 40% | 7.2 h | 14.4% | Productivity ratings +71%, satisfaction +52% |
| 3 | 60% | 10.8 h | 21.6% | Cooperation +55%, stress risk -57%; identified by the authors as optimal |
| 4 | 80% | 14.4 h | 28.8% | Micromanagement -74%, engagement +44%, communication transparency +65% |
| 5 | 100% | 18.0 h | 36.0% | Satisfaction, productivity, engagement and cooperation all decline |
The percentage effects are as reported by the study at each dosage; the study did not publish every metric at every level, so this table shows what was reported rather than a complete matrix. The hours columns are computed here from a baseline established by separate research, described below.
Read down that table and the strange thing is obvious. Benefits keep accumulating through four days, including the largest single effect in the whole dataset, a 74 percent drop in micromanagement. And then at five days everything collapses at once.
The authors’ explanation is social: meetings are how people maintain ties, and removing them entirely leaves remote workers isolated. That is almost certainly part of it. But it does not explain why the decline begins earlier, past a 60 percent reduction, which the authors also observed. For that you need to know how much of the meeting load was worth removing in the first place.
The arithmetic nobody ran
Steven Rogelberg, a professor of organisational science at the University of North Carolina at Charlotte, ran a survey of 632 employees across 20 industries that put numbers on the waste. Respondents reported an average of 18 hours a week in meetings against a median 50-hour week. They said they did not need to be in 30 percent of the meetings they attended, and could have skipped almost six of those 18 hours. The survey was conducted with a commercial partner rather than through peer review, so treat the dollar figures it generated with caution; the hours are the useful part and they are consistent with the meeting loads reported elsewhere.
Now put the two datasets together, which as far as we can find nobody has done.
The skippable pool is roughly 6.0 hours a week. A one-day policy removes 3.6 hours, comfortably inside it. A two-day policy removes 7.2 hours, which is already 1.2 hours more than the entire pool of meetings people said they did not need to attend. A three-day policy removes 10.8 hours, exceeding the waste pool by 4.8 hours. A five-day policy removes all 18.
Table 2. Meeting hours removed against the measured pool of skippable meeting time, CEOtudent editorial framework
| Dosage | Hours removed | Skippable pool | Necessary meeting hours displaced | What the policy is actually doing |
|---|---|---|---|---|
| 1 day | 3.6 h | 6.0 h | 0 h | Cutting waste only |
| 2 days | 7.2 h | 6.0 h | 1.2 h | Waste, plus a small amount of real coordination |
| 3 days | 10.8 h | 6.0 h | 4.8 h | Waste, plus substantial real coordination |
| 4 days | 14.4 h | 6.0 h | 8.4 h | Mostly relocating necessary coordination |
| 5 days | 18.0 h | 6.0 h | 12.0 h | Removing the coordination channel entirely |
This is the finding that should change how you read the policy. Past two days, a meeting-free-day policy is no longer a waste-reduction programme. It is a relocation programme. The necessary coordination does not disappear because you blocked the calendar; it moves into whatever channel you have left standing.
That reframes the entire question. Whether three or four meeting-free days help you has almost nothing to do with meetings and almost everything to do with the quality of the asynchronous channel absorbing the displaced hours. The study itself hints at this: at four days, the largest reported gain was that remaining communication became 65 percent more transparent, and the authors noted people checking written records instead of relying on recall. That is a description of a functioning written channel doing the coordination work.
It also explains the collapse at five days cleanly. At that point the entire 18 hours has been relocated and there is no synchronous channel at all, so the cases that genuinely need one, disagreement, ambiguity, anything emotionally loaded, have nowhere to go.
Which day you clear matters more than anyone says
Almost every organisation adopting this picks Friday, on the reasoning that Friday is already the quietest day and clearing it is the least disruptive.
That reasoning is exactly backwards if the goal is returning hours.
Microsoft’s Work Trend Index special report, published in June 2025 and drawing on both platform telemetry and a survey of 31,000 knowledge workers across 31 countries, reported the weekly distribution of meeting load. Tuesday is the heaviest day at 23 percent of the week’s meetings. Friday is the lightest at 16 percent. That leaves 61 percent spread across Monday, Wednesday and Thursday, an average of 20.3 percent each.
Apply those shares to the 18-hour baseline.
Table 3. Hours returned by which weekday you clear, CEOtudent editorial framework, combining published meeting-load distribution with a measured weekly meeting baseline
| Day cleared | Share of weekly meeting load | Hours returned per week | Relative to clearing Friday | Additional hours per year, 46 working weeks |
|---|---|---|---|---|
| Tuesday | 23.0% | 4.14 h | +43.8% | +58.0 h |
| Monday, Wednesday or Thursday | 20.3% each | 3.66 h | +27.1% | +35.9 h |
| Friday | 16.0% | 2.88 h | baseline | baseline |
Clearing Tuesday returns 43.8 percent more meeting hours than clearing Friday. Over a working year of 46 weeks that difference is 58 hours per person, roughly a week and a half of full-time work, for a policy that costs exactly the same to announce.
The obvious objection is that the heaviest day is heavy for a reason and those meetings will simply move. Some will. But the same logic applies to Friday’s meetings, and Friday has less to move. The counter-consideration worth taking seriously is that Tuesday’s load may be heavier precisely because it holds the week’s genuinely necessary synchronous work, in which case clearing it displaces more of the 1.2 to 4.8 necessary hours from Table 2 and raises the burden on your written channel. Which is the same conclusion again: the written channel is the binding constraint, not the calendar.
The failure mode the authors named
The study is explicit about how this goes wrong, and the mechanism is worth stating because it is the most common outcome.
If the policy is treated as a scheduling formality rather than a reduction, the meetings from the cleared day migrate to the remaining days. The week’s total load is unchanged, the surviving days become denser, and the result is a more compressed schedule and more stress than before. The authors describe cramming meetings into fewer days as solving nothing.
This is the diagnostic that separates a real policy from a cosmetic one. A month after adoption, measure the total weekly meeting hours per person, not the number of protected days. If total hours did not fall, the policy did not happen, whatever the calendar looks like.
What to do with this
The evidence supports a specific, bounded recommendation rather than an enthusiasm.
Start at two days, not three. Two days sits almost exactly at the edge of the measured waste pool, removing 7.2 hours against roughly 6.0 hours of meetings people say they did not need. That is the dosage with the most favourable ratio of benefit to relocated coordination, and it is where the largest reported productivity and satisfaction effects appeared.
Clear a heavy day rather than a light one. On published load distribution, Tuesday returns 43.8 percent more hours than Friday. If Tuesday is impossible, any of Monday, Wednesday or Thursday still returns 27.1 percent more than Friday.
Go to three days only after the written channel is real. Three days displaces 4.8 hours of coordination that people said they needed. If that lands in a channel where decisions are recorded and findable, the study’s cooperation and stress results are plausible. If it lands in an unstructured chat stream, you have removed a coordination mechanism and replaced it with noise.
Do not go to five. The study is unambiguous, and the arithmetic explains why.
Measure total meeting hours per person, monthly. It is the only number that distinguishes a reduction from a rearrangement.
Where this fits
The dosage question sits inside a larger one about how much synchronous time an organisation should hold. The true cost of meetings works out what that time is actually worth before you decide how much to cut, async-first work architecture is the written channel this policy depends on past two days, how AI meeting notes changed the meeting covers what happens to the record when attendance stops being the only way to know what was decided, and the deep work deficit is the measurement of what the fragmentation costs on the other side.
Frequently asked questions
Is the 71 percent productivity increase real?
It is a real finding and a widely misquoted one. It is a 71 percent rise in self-reported productivity ratings from pulse surveys at companies that adopted two meeting-free days, not a 71 percent rise in output. Treat it as strong evidence that people experience the change as a large improvement, and as no evidence at all about units produced.
How many meeting-free days should we start with?
Two. It is the most commonly adopted level in the study, chosen by 47 percent of the companies, it carries the largest reported productivity and satisfaction effects, and on the arithmetic above it sits at the boundary of the measured waste pool, so it removes almost entirely meetings people said they did not need.
Does this work for small companies?
Unknown from this evidence. Every company in the study had more than 1,000 employees. Small teams have a different meeting profile, typically fewer status meetings and a higher share of genuinely necessary coordination, which on the logic of Table 2 means the skippable pool is smaller and the same dosage displaces more necessary hours. The direction of the effect is plausible; the dosage is probably lower.
What about one-on-ones?
The study’s companies excluded even one-on-ones on meeting-free days. The reported benefits at three days included people replacing meetings with better-timed one-on-ones on the remaining days, so the finding is not that one-on-ones are waste; it is that concentrating them on meeting days worked better than scattering them.
Should we announce it as a wellbeing initiative?
Only if you also intend it as an operating change. The failure mode the authors identify is that meetings migrate to the remaining days and the week’s total load is unchanged, which produces a denser schedule and more stress while everyone believes a wellbeing measure has been introduced. That is worse than doing nothing, because it also spends the credibility you would need to try again.
Sources
- Laker, Pereira, Budhwar and Malik, The Surprising Impact of Meeting-Free Days, MIT Sloan Management Review, January 2022; accepted manuscript archived at the University of Reading Central Archive
- Rogelberg, survey of 632 employees across 20 industries on meeting load and necessity, University of North Carolina at Charlotte, 2022, commercially sponsored and not peer reviewed
- Microsoft WorkLab, Breaking Down the Infinite Workday, Work Trend Index special report, June 2025, based on platform telemetry and a survey of 31,000 knowledge workers in 31 countries conducted February to March 2025
- Rogelberg, The Surprising Science of Meetings: How You Can Lead Your Team to Peak Performance, Oxford University Press, 2019
- Allen and Rogelberg, Meetings at work: Perceived effectiveness and recommended improvements, Journal of Business Research, volume 68, 2015
This content was compiled with the support of AI following in-depth research, then written and prepared for publication by the CEOtudent editorial team.
This post is also available in:
