Structure is strategy

“The structure, not the plan, is your strategy.”

Stanley McChrystal

Team of teams

One of my favorite books from the last decade was Team of Teams. In it, General McChrystal documents how he figured out how to fight Al Qaeda. The US military had better plans but Al Qaeda had better structure - distributed, adaptive, each node able to make decisions independently because every node understood the mission. The plan became irrelevant once the first bullet was fired. The structure kept working. Effectively, everyone in Al Qaeda was #3.

“We killed about 20 of Al Qaeda’s ‘number threes’ over the past decade, but everyone in a network is a number three.”

Stanley McChrystal, Team of Teams

McChrystal’s task force was the opposite. World-class operators, world-class intel, world-class plans, strangled by a hierarchy built for an industrial-era enemy. They were faster than anyone alive, and still losing. The fight wasn’t about who had the better unit. It was about which side’s organization could process new information and act on it before the situation changed again.

What made Al Qaeda’s structure so durable was that no one had designed it. It had grown that way:

“… we saw no evidence that this inexplicable structure was the product of deliberate design; it seemed instead to have evolved through ongoing adaptation.”

So he rebuilt the org around two ideas:

  • Shared consciousness. Radical transparency. Every team sees the whole picture, in real time. The daily 7,000-person Operations & Intelligence call wasn’t a status meeting. It was the mechanism that made every node smart about the whole. You can’t push decisions down unless the people at the bottom actually have the context the people at the top used to have.
  • Empowered execution. Once everyone has the context, you let them act. Decision authority moves to where the information already is, not up the chain to where the rank is. McChrystal’s line: “It takes a network to defeat a network.”

The ranks didn’t go away. But the mental model of the org did:

“The task force still had ranks and each member was still assigned a particular team and sub-sub-command, but we all understood that we were now part of a network; when we visualised our own force on the whiteboards, it now took the form of webs and nodes, not tiers and silos.”

The shift in his own role was telling. He stopped trying to be the chess master moving pieces and started thinking of himself as a gardener, building the soil, the structures, and the rhythms that let the task force adapt on its own. Less commanding, more cultivating.

McChrystal talks through the idea in his own words here:

Lessons from the book

  • We live in a networked world, not a clockwork world. Taylorism is dead. Efficiency optimizes for a problem that isn’t the problem anymore. Adaptability is the problem.
  • Small teams are the core unit, because trust and shared context fit inside a small team. The hard problem is teams of teams: how independent teams stay coherent without a top-down plan holding them together.
  • Information flow is structure. The org chart matters less than the cadence and the meeting design. Who hears what, when, decides what’s actually executable.
  • Plans are perishable. Structure is durable. Invest in the structure.

It’s not unlike the over-used, over-studied, over-emulated Spotify model. Which I won’t describe here because it is so well documented.

What this means for your software product team

  • Org design - Keep teams small enough to self-organize. Structure determines what’s executable before strategy ever comes up.
  • Cadence - Daily = how you execute. Weekly = supervisory at best. Monthly = definitely supervisory. The structure of when people meet determines what gets attention.
  • The roadmap - It’s the structure for how the team understands its work. Not documentation. Architecture. If it’s a slide cranked out the night before the board meeting, the strategy is aspirational.

Last updated:

Next in Strategy All software execs have the same job →