The model and the machine

Picture a company on its first morning as itself. The carve-out has closed, the lawyers have gone home, the new name is on the building. The badges work. The payroll ran. Every system that could be listed on a schedule has been transferred, tested and signed for. The org chart is printed. The operating model sits in a binder on the new chief executive's desk, forty pages, every box filled.

By eleven o'clock a decision that used to take an afternoon has taken three days and is still not made.

Nothing is broken. That is the strange part. No system is down, no one has resigned, the people are the same people at the same desks, holding the titles they held a fortnight ago inside the parent. On paper the organisation is complete. In practice it has forgotten how to decide.

Here is what left the building without appearing on any schedule. The person three floors up who could be called and would simply know. The unwritten sense of which risk was worth taking, and to whom. The quiet map of whose judgment to trust when the written process ran out, which it always did. None of it was in the binder, because none of it could be. An operating model is a drawing of an organisation. The operating is what the drawing leaves out.

The discipline has known this for a long time, and said it most plainly from inside the houses that sell the models. In 1980 three McKinsey consultants titled a paper "Structure Is Not Organization" and argued that the picture of the thing is not the thing¹. Almost thirty years later Booz, the firm I came up in, put numbers to it. It surveyed more than a hundred thousand people across a thousand organisations and ranked what actually makes a strategy execute. Redrawing the chart came near the bottom. The two strongest levers were clarity about who owns which decision, and whether information moved². When a company stalls, the study noted, the first instinct is to reach for the structure, and it is usually the least useful thing you can touch.

The man who saw the informal side most clearly came up through the same houses. Jon Katzenbach spent a career on how work really gets done, on the culture and the networks and the ad hoc trust the chart never shows, and taught a generation to lead outside the lines³. He was right, and he was generous: he kept trying to give leaders something they could hold, a method, a critical few behaviours, a way in. That generosity is the honest tell. Even the discipline's clearest witness to the un-drawable reached, in the end, for something to draw. The operating model is only the latest and fullest of those reachings, twelve elements now, culture and decision rights and behaviour each given a box of its own⁴. Every version more faithful, more nearly alive, and every one still a picture of an organisation rather than the living thing a picture cannot hold.

The variable sits one level up from the quality of the model. It is the share of the operating that was ever in any model at all, and that share has mostly been studied at rest, as a thing to harness. Less has been said about what happens when it is destroyed. And it does get destroyed, in an instant, on the morning a discontinuity removes the conditions that grew it. A new chief executive. A carve-out. A merger above all. The people who did this work for a living, the ones who sat inside the integrations, knew the secret the decks never printed: combine two companies and the charts merge on day one, while the informal organisations do not merge for years, if they ever do. Everything that can be drawn lines up. Everything that cannot goes dark. The formal organisation survives the event intact. The informal one, the one actually running the company, simply stops. And because keeping it alive was never anyone's job, no one notices it has gone until the decisions start taking three days.

This is why the work that matters in these moments looks like nothing on an org chart. It is not the design of a new model. The new model is already printed, complete, sitting in the binder. It is the slow rebuilding of the undrawn thing: the trust, the judgment, the quiet map of who to call, reassembled from the inside by people who stay in the room long enough to grow it back before business as usual is allowed to resume.

Back in the building, the lights are on and the model is immaculate. Every box is filled and nothing is moving. Somewhere in the quiet an organisation is trying to remember the one thing it was never able to write down. The model is not the machine. The machine was always the part you could not draw.

Notes

1. Robert H. Waterman Jr., Thomas J. Peters and Julien R. Phillips, "Structure Is Not Organization," Business Horizons, vol. 23, no. 3 (June 1980), pp. 14–26. The source of the line that the picture of the thing is not the thing, and the paper that introduced the McKinsey 7S framework. ↩
2. Gary L. Neilson, Karla L. Martin and Elizabeth Powers, "The Secrets to Successful Strategy Execution," Harvard Business Review (June 2008). Booz & Company's analysis of survey responses from more than 125,000 people across roughly 1,000 organisations in over 50 countries. Of the four building blocks of execution, decision rights and information flow ranked as the strongest levers and structural change among the weakest. ↩
3. Jon R. Katzenbach and Zia Khan, Leading Outside the Lines: How to Mobilize the Informal Organization (2010). The formal and informal distinction runs through Katzenbach's work from The Wisdom of Teams (with Douglas Smith, 1992) to the Katzenbach Center's culture methodology in The Critical Few (2018). ↩
4. McKinsey & Company, "What is an operating model?" (McKinsey Explainers, November 2025). The contemporary twelve-element, system-not-structure formulation. ↩