Organizational Flow — Appendix
Principles
Not rules — observations, developed across the chapters of this paper and gathered here for reference. Each one is argued for in full in the chapter it's tagged to; this page is the index, not a replacement for reading them.
Part II · A Meeting Grew Here
- Purpose gives direction.
- Focus is what creates speed. Purpose is what creates focus.
- Purpose gives a person a reason. Conditions let that reason turn into something that stands.
Part I · What is Organizational Flow?
- Technology is rarely the bottleneck. Friction is.
- Flow over friction.
- Flow is an organizational capability.
- Flow has a human dependency.
- You cannot run a machine organization and expect it to produce living things.
Part I · The Cost of Value Never Created
- Friction never announces itself as failure. It arrives as everyone being busy.
- Friction lives between people's work, which is why it belongs to no one.
- Remove friction before adding more.
- Everything is either flow or friction.
Part II · A Meeting Grew Here
- Ownership before governance.
- The goal is not maximum autonomy or maximum control. It is the minimum structure needed to preserve coherence while value is created with as little friction as possible.
- Share what is genuinely common. Separate where the requirements meaningfully differ.
- Governance grows where ownership is unclear.
- Ownership isn't about control. It's about accountability.
- Ownership creates progress when it's taken.
- Clear ownership is one of the strongest predictors of an organization's ability to adapt.
- Ownership is knowing, without asking, what's yours to decide.
- A control that people avoid isn't a control.
Part II · The Wait Nobody Logged
- Contribution creates flow.
- People create the conditions for each other's success.
- Trust grows when people consistently help others move.
- Organizations are different. Interdependence isn't.
- People own. AI amplifies. Organizations learn.
- AI should amplify capability — not replace accountability.
Part II · It Breaks in the Joints
- Information should move, not accumulate.
- Information creates value when it's shared.
- Information should be captured once and used many times.
- Copying and reconciling information creates friction.
- The question is never whether we have the information. It is whose decision it reaches, and when.
Part II · It Breaks in the Joints
- Decisions create momentum.
- Slow decisions slow organizations.
- A good decision made early beats a perfect decision made too late.
- Move the decision to the information, not the information to the decision.
- Delay is rarely caused by the difficulty of the decision. It is caused by uncertainty about who makes it.
Part II · It Breaks in the Joints
- Execution is where strategy learns.
- Strategy is a hypothesis until it's executed.
- Execution creates learning.
- Flow breaks at the joints, not inside the stages.
Part II · We Have Solved This Before
- Experience is automatic. Learning is a choice.
- Organizations learn through execution.
- Learning compounds over time.
- Learning that produces an artefact has an obvious place to stop.
- An organization that doesn't learn isn't slow. It's starting over, continuously.
Part II · What Survives the Reorg
- Capabilities endure.
- Technology enables capabilities. It doesn't define them.
- Organizations create value through capabilities — not technology.
Part II · The Long Way to a Yes
- Architecture should make the organization easier to change.
- Architecture creates the conditions for adaptation.
- Architecture exists to improve flow.
- Architecture is cultivation.
- Architecture is about conditions, not control.
- Architecture is a competence, not a role.
- Most bad architecture is not wrong so much as premature.
Part II · The Team That Has to Ask
- Draw the team boundary around the capability.
- Size a team by cognitive load, not by headcount.
- Collaboration is a handover you have chosen to keep. A service is one you have removed.
- Representation is the problem, not the solution.
- A team you have to interrupt someone else to run is a queue with a name.
Part II · You Have to Know a Guy
- An interface nobody can find isn't an interface. It's a person you have to know.
- Automate it, then make it self-service, then run a proper queue — and only then, a person you have to know.
- A form is not self-service. If the request lands in somebody's list, it is a queue.
- Undocumented interfaces become complicated because they were undocumented.
- Publish internally as though you were publishing externally.
- Synchronous couples availability. Asynchronous trades certainty for independence.
- Asking someone who knows is the most reliable friction generator in any organization.
Part III · Settling the Few Things
- Do what is best for the company.
- Guardrails are a short list of decisions everyone else no longer has to make.
- Standardise where variation costs more than it buys.
- A guardrail that makes the organization harder to change has failed.
- Making the decision is the job. Writing it down takes ten minutes.
- A decision record nobody can find is the same as no decision record.
Part II · Decided Upstairs, Felt Downstairs
- Leaders create conditions. People create flow.
- Expectations create responsibility.
- Leadership is about cultivating, not controlling.
- Reorganizations are architecture, whether or not anyone reviews them as design.
- Fund capabilities, not only projects.
Part II · Faster, and Still Wrong
- Technology scales capability. People scale judgement.
- Technology multiplies whatever is already there. It doesn't repair it.
- Every tool encodes an assumption about who decides what.
- AI amplifies organizations. It does not repair them.
- AI amplifies human skills — it does not replace human accountability.
Part III · Growing the Team
- Cultivate, don't control.
- Organizations should be designed to learn.
- Organizations are living systems, not machines.
- Change is a capability, not a project.
- Whatever an organization builds takes the shape of the organization that built it.
- A team of only builders can build anything and decide nothing.
- Distribute the competence and the guardrails together.