Back to blog
9 March 2026/9 min read

Your Meetings Aren't the Problem, Your Teams Are

Oversized teams create a coordination tax that AI has made dramatically more expensive, making smaller high-trust teams the real operating advantage.

Your Meetings Aren't the Problem, Your Teams Are

Two years ago I wrote on here that 2023 would be remembered as the year AI went mainstream. I was right about that. What I got wrong was assuming the biggest disruption would be technological.

It's not. It's organisational.

I have spent 15 years delivering technology transformation across financial services, government and enterprise. I have an MBA in Finance. I am currently completing a Graduate Certificate in Applied AI. And after all of that, the single biggest pattern I see killing value in Australian organisations right now has nothing to do with models, tooling or strategy decks.

It is team size.

Your meetings problem is not a meetings problem. It is a coordination tax that scales with every person you add to a team, and AI just made that tax an order of magnitude more expensive.

The maths is not new. Dunbar established in 1992 that the human brain sustains deep coordination with about five people. The military tested it empirically. Bezos arrived at the same number with the two pizza rule. Fred Brooks proved in 1975 that adding engineers to a software project makes it slower, not faster. Three entirely separate disciplines, evolutionary psychology, military operations and software engineering, all landed on the same answer.

What AI changed is the consequence of ignoring that answer.

Before AI, a five person team produced X output. Adding a sixth gave you diminishing returns because coordination overhead grew faster than capacity. Annoying, but manageable when each person generated $250,000 a year in value. An effective AI native 5 person team can produce 5 to 10x more, output is no longer the constraint, coordination is. The revenue per employee data from AI native companies is staggering. Lovable, Midjourney, ElevenLabs, all running 5x, 10x or even 25x multiples above traditional SaaS benchmarks with tiny teams.

Each person on your team is now capable of generating millions in value, if you know how to use them, this means the coordination cost of person number six is no longer a minor tax. It is measured in millions of lost productivity. Every meeting on your calendar exists because someone decided coordination was worth the cost. At $250,000 per person, it often was. At $2 million per person, most of those meetings are net negative.

And this is not just a SaaS story. If you are running enterprise delivery, whether that is technology transformation, infrastructure programs or capability builds across government and financial services, the same maths applies. The difference is you feel it in delivery cost blowouts and capability timelines that stretch from months into years. Every oversized team, every unnecessary coordination layer, every alignment meeting that produces another alignment meeting is time and money bleeding out of programs that were already under pressure. AI did not create that problem, but it made the cost of tolerating it indefensible.

I have seen this firsthand more times than I can count. Delivery teams of 30 plus people spread across multiple organisations, all mobilised to deliver what are fundamentally simple D365 implementations. The team is so large that the overwhelming majority of effort goes into planning, governance, status reporting and cross-stream alignment rather than actually building features. The work that should take months takes years. The budget doubles. And inevitably the project lands with reduced scope and diluted benefits, delivering less than what a focused team of five to eight people could have shipped in a quarter of the time.

Conservatively, the coordination overhead alone inflates delivery costs by at least 30%. That is what I have seen consistently across 15 years of enterprise delivery, and on larger programs with multiple vendors, cross-agency governance and layered approval chains, that number climbs well above 30%. I have watched programs where more than half the budget was consumed by the machinery of coordination before a single feature was delivered. On a small engagement that is hundreds of thousands of dollars burned on alignment instead of outcomes. On a large implementation it runs into the millions. Money that was budgeted for capability development spent instead on making sure 30 people across four organisations all understand what the other 29 are doing.

And to be clear, this is not a criticism of the project managers. They are usually doing exactly what they should be doing, working diligently within the prescribed frameworks and governance structures they have been given. The problem is not the people executing the process. The problem is that the structure itself demands a level of coordination that consumes the capacity it was supposed to protect. The PM is not failing. The operating model is failing them.

The irony is brutal. The size of the team, which was supposed to derisk the program, is the single biggest risk to its success. I have watched this pattern repeat across financial services, government and professional services and it plays out almost identically every time.

And here is a question I think every executive needs to sit with honestly.

How many of you actually know the value your teams produce? Not the cost. You know the cost. Every leader can tell you what their team costs in wages, benefits and overhead down to the dollar. But can you tell me what that team delivers in value? Can you quantify the output, the revenue generated, the problems solved, the decisions made well?

Most cannot. And that gap is where the real damage hides.

If you do not know the value your team is producing, you have no way of knowing what meetings, admin and coordination overhead are actually costing you in lost delivery. You see a calendar full of syncs and standups and you think "that is just how work gets done." But what you are really looking at is an invisible tax on your most valuable people's output, and you cannot even measure it because you never measured the output in the first place.

There is an older principle that makes this even more uncomfortable. The Pareto Principle, the 80/20 rule, has been observed across industries for decades: roughly 20% of a team produces around 80% of the meaningful output. In most large delivery teams, the real work is being carried by a small core of high performers while the remaining coordination structure exists largely to keep everyone else aligned. That is not a people problem. It is a structural one. You have built a machine where the majority of effort services the machine itself rather than the mission.

What high performing digital teams demonstrate is that the 80/20 split is not a law of nature. It is a symptom of poor team design. When you compress a team to five or six people with clear ownership, full context and the right tooling, those numbers shift dramatically. You stop carrying coordination passengers and start operating with a team where every person is a meaningful contributor. The best teams I have worked with do not beat the 80/20 rule by finding better people. They beat it by removing the structure that made 80% of the team's energy disappear into overhead in the first place.

The companies getting this right are obsessive about value per person, not cost per person. That is the shift.

And here is where I think most executives are getting it catastrophically wrong.

The conversation I keep hearing is about efficiency. Same mission, fewer bodies, lower burn. That is a staggering failure of imagination. If you have 500 people and each just became 5 to 10 times more capable, the correct response is not "I can run my company with 50." The correct response is "I now have the productive capacity of 3,000 people. What was I previously unable to do?"

You did not get a cost reduction. You got a force multiplier.

The companies defining this era are not shrinking to protect margins on their current ambition. They are reorganising into small, high trust teams of five and pointing them at missions that were impossible 18 months ago. Scouts (solo operators with full AI tooling) explore new territory fast. Strike teams of five execute where correctness matters, because in a world where AI made volume free, correctness is the only scarce resource left.

A Harvard Business School study published this year tested it directly. 776 professionals at Procter and Gamble on real innovation challenges. Teams using AI were three times more likely to produce ideas in the top 10% of quality. Not three times more output. Three times more likely to be right. AI also broke functional silos, extending each person's competence into adjacent domains. That is the mechanism. Five excellent generalists using AI can cover territory that previously required a department of specialists.

I see this every day building products across my own ventures. The old model of staffing up and coordinating across layers of management is not just inefficient now. It is structurally obsolete for any team trying to move at the speed the market demands.

So if your calendar next week is wall to wall syncs, standups and alignment sessions, I would gently suggest the problem is not your meeting culture. It is that your teams are three to 10 times too big for an era where every additional person on a team costs you time and money. Stand up newer, focused teams, focused on solving context constrained problems in isolation or with loose networks among other teams.

The leaders who restructure around this will build the defining companies of the next decade. The ones who do not will spend a lot of time in very well documented meetings, wondering why companies a fraction of their size are eating their lunch.

It is not time to shrink your headcount. It is time to shrink your teams. Focus on rapid iteration and capability development. Put your best people in groups small enough to actually trust each other, arm them with AI, and point them at something worth building.

How many people are on your largest delivery team right now? And honestly, what would happen if you halved that number tomorrow?

References