Your team does not need fewer meetings. It needs better designed ones.

A recent calendar review showed the familiar pattern. Twenty-three meetings the previous week. One standing weekly had already been cancelled. Two recurring syncs had already been declined. The team still felt slower than three months earlier.

When we looked at what had actually moved, the answer was uncomfortable. None of the important decisions had closed. Three minor priorities had shifted. The big calls, the architecture path, the next quarter's portfolio, the rework decision on a stalled feature, were still open. Some had been open for six weeks.

The problem was never the meeting count. The problem was that none of those meetings had been designed to make a decision.

What the study says

The dominant advice in delivery management for the last few years has been simple. Cut meetings, protect focus time, ship more. It is correct enough to be repeated, and wrong enough to mislead.

A 2025 field study by Alexander Romney, Joseph Allen and Zahra Heydarifard, published in Business Horizons, surveyed 199 full-time employees on the relationship between meeting load and three outcomes: active participation, work engagement, and creative performance. The mean was 7 meetings per week, with a standard deviation of nearly 6. They tested whether the effect was linear or curvilinear.

It is curvilinear. An inverted U. As meeting load rises, creative performance and active participation both rise until a threshold, after which both drop. The quadratic term is significant for creative performance (B = -0.537, p = 0.004) and for participation (B = -0.381, p = 0.025). Engagement is more stubborn. The curve is suggested but not statistically supported (p = 0.116).

Two moderators are worth flagging. Virtual format sharpens the curve on creativity, with the peak coming earlier and the decline steeper. Conscientiousness flattens the curve on participation, so orderly people absorb more meetings before they break.

Caveats first. This is cross-sectional, self-report, on a Prolific sample. "Creative performance" is not "cycle time." The authors do not name a specific meeting-count threshold, and I will not invent one. What they do show, with peer-reviewed regression coefficients rather than a LinkedIn opinion, is that the function is curved. Not monotonic.

The practical implication is not "have fewer meetings." It is that the meetings on the falling side of the curve are doing something different from the meetings on the rising side. If you do not separate the two, calendar surgery will stay blunt.

What actually breaks first

In the same paper, the authors point at a mechanism most teams recognise but rarely name. A meeting that fails to settle what it was meant to settle tends to generate further meetings to clean up the gap. The growth in meeting load is partly endogenous. Bad meetings create more meetings. The calendar is not the disease. It is the symptom.

What is a "bad meeting" in this sense? The authors are specific. A meeting without a clear purpose. A meeting that requires no genuine collaboration but is held anyway because someone wanted to share information. A meeting that ends without identified takeaways and assignments. They also cite figures suggesting that in some organisations only a small share of electronic meeting invitations carry an identifiable purpose at all.

This is the part Delivery Managers and Engineering Managers should sit with. If your standing syncs end without a single named decision and a single named owner, you are not "aligning." You are accumulating a queue of latent decisions that will surface later, more expensively, in incidents, in rework, in stakeholder escalations, or in the next planning meeting where the same topic appears again.

Donald Reinertsen made the economic version of this point fifteen years ago. The unit of waste in product development is rarely an idle developer. It is a decision sitting in a queue. Principles of Product Development Flow keeps coming back to one rule. If you measure one thing, measure cost of delay. A decision delayed by six weeks does not become a six-week problem when it is finally taken. It compounds. Every choice downstream that was made without that input was made on a guess.

When a team complains about "too many meetings," the evidence suggests a more precise prior. Too few of those meetings produce decisions, and the decisions they fail to produce are themselves loading the calendar.

Two low-effort design moves the evidence supports

Two interventions show up in both the Romney paper and in current practitioner writing. Neither is new. Both are routinely skipped.

Silent start. Romney and colleagues call it "the silent meeting." The group spends the first few minutes reading a short brief, or writing for one minute, before anyone speaks. Barry Overeem at The Liberators writes about the same mechanism through Liberating Structures, with a minute of individual silent reflection before opening 1-2-4-ALL or Impromptu Networking. His observation, after years of running these formats, is that without that minute "mainly quick thinkers get space to speak, while ideas from slower thinkers are lost."

This is not just a facilitation preference. In a meeting that is supposed to decide something, "ideas from slower thinkers are lost" means the input set is biased before the decision even starts. That is how decisions get reopened later. The minute of silence is not friction. It is cheap decision-quality control.

The purpose-and-collaboration filter. Before scheduling a meeting, two questions. Does this meeting have a defined purpose? Does that purpose actually require collaboration? Romney and colleagues argue that if either answer is no, write a memo instead. Record an audio note. Drop it in the channel. The most common meeting type in their data is the information-sharing meeting, and information-sharing is the textbook case of work that does not need synchronous presence.

Apply only those two rules. Silent start on decision meetings, memo-or-skip on information meetings. The calendar will not become perfect, but it will start to show which meetings deserve synchronous time and which ones were only habits.

A practical audit

Sample the last 10 calendar invites the Delivery Manager accepted. For each one, ask. Was there a stated purpose? Did the meeting produce a named decision with an owner? Did the next meeting on the same topic reopen the same question? It takes about forty minutes and is more honest than asking the team how meetings feel.

Three signals are worth looking for. The same topic surfacing in three or more consecutive standings without a named decision. Information-share meetings whose attendees are managers, not the people who need the information. Invitations with no agenda where the only owner is the leader's calendar assistant. Each one is a queue dressed as a meeting.

A useful test for the team. Mark, for the past two weeks, every meeting where you can later point to a specific decision that came out of it. If the number is low, treat that as a meeting-design problem before treating it as a meeting-overload problem. The inverted U is the empirical reason "cut meetings" advice will keep failing teams that skip this distinction.

What this is not

Not an argument for more meetings. The downward leg of the inverted U is real, and many teams sit on it. If your team is past the peak, fewer meetings will help.

Not an argument that silent starts and purpose filters are sufficient. Decision latency has structural drivers that no facilitation pattern fixes. Unclear decision rights, missing data, organisational fear of being wrong on record. Those need separate work.

The better conversation is this. Stop arguing about meeting count first. Start by asking which meetings make decisions, and design those properly.

Sources

• Overeem, B. (2025, December 29). Designing better conversations starts with silence. The Liberators. medium.com/the-liberators
• Overeem, B. (2026, January 11). Crappy conversations. The Liberators. medium.com/the-liberators
• Reinertsen, D. G. (2009). The principles of product development flow: Second generation lean product development. Celeritas Publishing. search.worldcat.org/title/435994279
• Romney, A. C., Allen, J. A., & Heydarifard, Z. (2025). Meeting load paradox: Balancing the benefits and burdens of work meetings. Business Horizons, 68(1), 33–43. doi.org/10.1016/j.bushor.2023.10.002