Prologue
Why now?
This essay and the pod structure it contains is a response to a specific technological moment highlighted by Jack Dorsey and Roelof Botha in their essay hierarchy to intelligence.
With the aims of speed and accountability, CoTreat’s pod structure is built around three key convictions.
- AI is the new communication and coordination layer.
- Small teams perform better than individuals or large teams.
- Small teams need to be set up in a specific way to enable great work.
A pod is most often a group of three to six people, formed around a single measurable mission, within a defined timeframe. Diversity of expertise is essential. Role clarity and accountability is essential. Some pods last weeks. Some last months. None, except perhaps the CEO’s central pod, lasts longer than that.
AI is the new communication and coordination layer.
Small teams perform better than individuals or large teams.
Small teams need to be set up in a specific way to enable great work.
I.
Mission, metrics and incentives
A pod begins with a mission. One sentence, usually partially qualitative, partially quantitative that names what the pod should deliver and by when. In bygone times, this might have been "capture this watchtower by winter solstice" or "guard this flock of sheep till full moon."
The mission also has clear upstream metrics to ensure the entire team knows that a pod is headed in the right direction.
The pod will also have clear rewards agreed for achieving the mission. This was a big omission from my original memo.
A pod’s lifespan
Mission set
One sentence. What, and by when.
Ships autonomously
No external bottleneck allowed.
4–8 weeks
Long enough to be real. Short enough to stay sharp.
Dissolves
Mission met or not — inertia is not an option.
II.
Autonomous
A pod is autonomous when it can ship independently. If a pod cannot ship autonomously, and becomes bottlenecked by an individual or pod external to it, it is not a pod and must be dissolved immediately.
III.
Impermanent
Pods are impermanent by design. The default lifespan is short. 4-8 weeks typically. Long enough to deliver something real. Short enough that nobody settles in. Achieve the mission or not, the pod will dissolve to prevent inertia and ossification.
IV.
Hats
Inside every pod there are three structural hats. These are taken directly from Jack Dorsey’s hierarchy to intelligence.
Structural hats
One hat, one pod, at a time - that is the discipline.
The Directly Responsible Individual. One person, accountable per pod.
The Individual Contributor can be any role. Usually picked by the DRI during pod formation to help achieve their mission.
The Player-Coach teaches by doing and shows what great looks like.
I’m increasingly of the opinion that most people should only wear one hat, and be in one pod at a time.
V.
Diversity of Expertise
Every pod must contain three different angles of expertise.
- Not three engineers.
- Not 2 account executives and a SDR.
- Three people with genuinely different ways of looking at the mission.
This is the diversity principle in its purest sense.
VI.
Reinforcements
Pods begin critically lean.
Most new pods have up to three people with a player coached assigned. The ideal construct is one operator per angle, three hats distributed across them.
Reinforcements are assigned only when the pod is delivering and the metrics are moving. One or all of the expertise angles are reinforced and, the pod grows from 3 to up to 6 people.
The second in commands also make excellent seeds for future pods through the intense learning and organic experience of a winning pod.
Reinforcement, earned
Deep Customer
Deep Technical
Amplifier-Synthesiser
Three people. One operator per angle, three hats distributed across them.
VII.
CGAI
The pod structure would not have worked a year ago.
Pods dissolve and reform. Nobody sits permanently anywhere. How does the company not descend into total chaos?
The answer in part is CGAI. Company General AI. The connective tissue that lets the rest of the structure work.
CGAI has 7 angles.
CGAI’s seven angles
Hover a point to see what flows through it.
Everything relevant from the outside world is ingested: customers, leads, investors, partners, competitors, and much more.
Everything visible inside the company flows in automatically. Slack threads, Notion pages, Jira tickets, Hubspot records, Google Drive documents, GitHub commits, every public artefact the company produces is routed into CGAI without anyone having to think about it.
Operators push their personal updates. Team members are also encouraged and expected to push valuable information siloed in their heads to CGAI directly. End of day blogs or simple voice memos.
Permissions are paramount. Not everyone in the company should see everything. Rigerous permissions, governance frameworks created and continuously tested are paramount to CGAI longevity.
CGAI pushes, not just pulls. A middle managers job was to actively inform, align and inspire. CGAI aims to do the same.
Every pod is connected directly to the CEOs Central Pod via CGAI, with nothing in between.
If we execute this well internally, CGAI becomes part of our product offering. Every business small and large will need AI at the centre to win the future.
Epilogue
Shoulders of Giants
This model is not invented from nothing. At least three ancestors deserve naming.
The closest is the Three Amigos practice from agile software development. The Three Amigos puts a developer, a tester, and a business or product representative together, and they work end-end.
Another ancestor is the Spotify guild model, popularised in the early 2010s: squads, tribes, chapters, and guilds.
The most recent and direct ancestor is of course Block. In late March 2026, Jack Dorsey and Roelof Botha published From Hierarchy to Intelligence.
Lastly, the right test of this model is whether the entire company moves faster than in any other arrangement. And whether the people inside pods are happy, and doing the best work of their careers.