Part III — Cultivating Organizational Flow · Chapter 36 · 4 min read · First Public Draft

Cultivation

Tending What Grew

You cannot redesign a structure nobody designed. You can prune one.

Reduces frictionCreates valueOwnership clarity

Two observations in Part II end without anywhere obvious to go. It Grew Like That says the shape arrived one reasonable decision at a time and has no author to consult. The Conditions Don’t Come in the Box says the borrowed shape will not bring what made it work.

Between them they rule out the two things organizations normally do: redesign it, or copy somebody else’s. Which leaves the question this chapter is for — what do you actually do with a structure that grew?

You tend it. Which is slower, less impressive, and the only thing I have seen work.

Start with what it makes hard, not with what it is

The instinct is to map the current structure and look for what is wrong with it. That produces an accurate diagram and no decisions, because there is nothing wrong with it — every part of it made sense to somebody at some point.

The more useful question is narrower. Take the three things that most recently took far longer than they should have, and for each one, trace where it waited. Not who was slow. Where it sat.

You will get a small number of places, repeatedly. That set is your actual structure — the load-bearing shape, as opposed to the drawn one — and it is usually four or five boundaries rather than the forty on the chart.

This works because friction is honest in a way org charts are not. The chart tells you what somebody intended. The waiting tells you what exists.

Prune one thing, then wait

The strong temptation, once you can see it, is to fix all five at once and call it a reorganization. Which converts tending back into redesign, and reintroduces every problem Part II described — including that nobody afterwards will be able to say why the new shape is the way it is.

One at a time is not caution, it is information. A grown structure has dependencies nobody has documented, and the only reliable way to find them is to change one thing and see what complains. Change five and you learn nothing about any of them.

Wait long enough to see the second-order effect, which is usually a quarter. Things get briefly worse first — that is what removing a compensation feels like from inside, and it is the point at which most attempts get reversed.

Expect less resistance than you have budgeted for

This is the part that surprises people, and it follows directly from the observation.

Nobody chose the current arrangement, so nobody is defending it. There is no author whose reputation is attached, no original decision anyone is protecting. What looks like resistance is nearly always something narrower: a person who has built a working accommodation around the awkward boundary and is not confident the replacement will work as well.

That is a solvable conversation, and it is a completely different conversation from the one people prepare for. Most of the political energy spent on structural change is spent arguing with an opponent who is not there.

Borrow questions, never shapes

The Conditions Don’t Come in the Box argues that the transferable part is the question somebody asked, not the structure it produced.

So when something works somewhere else, the useful exercise is archaeological: what problem was this solving, and do we have that problem? Usually the answer is a version of it, differently shaped — which is far more useful than the model, because it sends you looking at your own conditions rather than at their diagram.

The failure mode to watch for in yourself is the one I know best. You come back with something good, and you want to give people the answer rather than the question, because the answer is the part that took you three days at a conference and the question sounds too simple to be worth flying anywhere for.

Keep a record of why, because the next generation won’t have one

Everything above produces a structure that grew — just more deliberately. Which will be equally opaque in eight years unless somebody writes down not the shape, but the reason.

One paragraph per boundary change: what was waiting, what we moved, what we expected. That is enough for whoever inherits it to know whether the reason still holds, which is the only question that matters and the one no org chart has ever answered.

A structure nobody can explain will be reorganized eventually, by somebody who cannot see what it was for. Writing down the why is the cheapest defence against your own successor.

This cultivates: It Grew Like That and The Conditions Don’t Come in the Box, in Part II.