Part IV — Applying Organizational Flow · Chapter 59 · 6 min read · First Public Draft
Working This Way
What it actually looks like to run an organization through this lens, week to week.
Jump to a chapter
Narrated with Microsoft's neural voice, not me — a recording.
If this paper has done its job, it hasn’t given you another framework to implement.
It has given you another way to see — and a way of working that follows from it, which is the part worth being concrete about, because a lens you cannot act through is just a nice sentence.
So this chapter is about the practice. What changes in an ordinary week, from the two vantage points this paper is most often read from — the one that sets the conditions, and the one that works on them directly.
If you run the organization
Four habits, and none of them require a mandate, a programme, or anyone’s permission.
Ask where it waited, not why it was late. The standard executive question — why is this taking so long — reliably produces a defence, because it is heard as an accusation about effort. The same information comes out of a different question: between the moment this was decided and the moment it reached a customer, where did it sit still, and for how long? That question is answerable, nobody has to protect themselves from it, and the answer is almost always somewhere nobody expected.
Fix ownership before you fix process. When something is going badly, the available moves are usually a new forum, a new process, or a new report. Before any of those, establish whether one person can say I own this and I decided. If they can’t, every process you add is a compensation, and you will be maintaining it for years.
Count what you removed. Most executive scorecards only go up. Keep a private list of the meetings that stopped happening, the approvals that were retired, the decisions that no longer come to you. If that list is empty after two quarters, nothing structural has changed, whatever else has improved.
Protect one channel to the outside that nobody can optimize. Someone who talks to real customers often enough to contradict the dashboard. This is the first thing cut when the quarter gets tight, and it is the only defence against becoming extremely efficient at the wrong thing.
If you work on the organization’s ability to change
This is a competence rather than a job title, and the people holding it are not always called architects. They are operating-model people, transformation leads, engineering and product managers, chiefs of staff — anyone whose actual subject is how the organization decides and how easily it can change its mind. Architects are among them, and the work described here is not the one on most architecture job descriptions, which is worth stating plainly.
Spend your time on boundaries, not on artifacts. A diagram, a target operating model, a strategy deck — each is a way of thinking, and occasionally a way of explaining. None of them is the work. The work is where the lines fall, who is allowed to decide what, and which capability is currently split across two departments that both believe the other owns it.
Turn up where the architectural decisions are actually made. They are made in budget rounds, reorganizations, hiring decisions and vendor selections — mostly by people who were told architecture was someone else’s job. Anyone responsible for the organization’s ability to change has to be present in those rooms, not only in the technical forums.
Settle a guardrail rather than review a decision. Every review you perform is a decision that had to travel. Every guardrail you settle removes a whole class of them permanently. Reviewing is more visible and more immediately satisfying, which is why most architecture functions do too much of it.
Say what you would need to see to be wrong. Arguments about structure become religious wars precisely when nobody has stated what evidence would settle them. Volunteering the conditions under which you would change your mind is the fastest way to be believed, and it costs nothing you were actually using.
Measure something, even badly. Four hours of handoff loss, self-reported by one team, is worth more in a leadership conversation than any model you will ever draw — because it is arithmetic, and arithmetic can be argued with.
If you sit in the middle
Both sections above are written for people with a mandate. The person most trapped in the friction has neither, which is why this section exists.
Middle management is the only place in an organization where the paradox is fully loaded. You are rewarded on headcount, in most places still expected to be the person who decides, and structurally required to route decisions upward through people who hold less detail about them than whoever raised them. Every one of those is somebody else’s design, and all three land on you.
It is also, and this is the part that gets missed, the only position from which decision authority can actually be handed downward. An executive can announce that teams are empowered. Only the person the decision currently routes through can stop being in the route.
Push one decision down and hold your nerve when it goes badly. The same move as elsewhere in the paper, and considerably harder here, because you will be asked about the outcome by somebody who does not know you delegated it. Answering that was the team’s call and I stand behind it is the whole practice, in one sentence, and it is expensive exactly once.
Stop being the translation layer. Much of what reaches you is a question that would be answered better by connecting two people directly. Every time you relay instead of introduce, you have made yourself load-bearing for a conversation that did not need you — and No Incident, No Story explains why nobody will notice you stopped.
Say what the structure makes hard, upward, before it is announced. You can see the three things a reorganization will slow down, and the room deciding it usually cannot. That information has a short window: raised beforehand it is design input, raised afterwards it is resistance. Same sentence, different reception.
And protect your own list. You are also the place where waiting accumulates, because everything routes through you. Ten Minutes Today applies to you more than to anyone else in the organization, and the answer to who is waiting on me is longer here than it is anywhere else.
None of that is a complaint about managers, and none of it is sympathy. It is a description of the one seat that can hand authority downward, held by people who are usually measured on holding onto it.
What all three vantage points share
The discipline underneath all three is patience with things that do not report well.
Conditions are slow, mostly invisible, and their evidence is an absence — a meeting that no longer happens, a wait nobody experienced. There is no way to make that legible on a quarterly slide without turning it into something else, which is exactly the failure mode the previous chapter describes.
So the work is largely unwitnessed, and the people doing it get very little credit, and it compounds anyway. That is the trade. An organization capable of valuing it has already solved most of what this paper describes.