Innovation6 min readFebruary 26, 2026

Designing organizations for continuous change

Every reorganisation announced as a bold new operating model is also a quiet admission: the structure that preceded it could not absorb what happened to it. The most adaptable companies are not the ones that reorganise most decisively. They are the ones that rarely need to, because change is a property of how they are built rather than an event they have to schedule.

66%: Of structural changes absorbed without a central programme, by Q7
3.1×: Longer recovery after a scheduled reorg than after a continuous adjustment
19 mo: Median interval between reorganisations in the groups we measured

The reorganisation is the symptom

The big reorg has a familiar shape. A chart is redrawn over the summer, announced in October, and takes effect in January. Six months of ambiguity follow about who owns what, and by the time the new structure reaches full productivity the conditions that justified it have changed again. Across the European groups we work with, the median interval between reorganisations is nineteen months — shorter than the time most of them need to make a single one work. The organisation is never in its designed state. It lives permanently inside a transition it keeps describing as temporary.

It is tempting to read this as poor execution. The trigger is almost always legitimate: a new customer segment, a regulatory change, an acquisition that arrives with its own logic. The failure sits upstream. The existing structure had no way to answer the question without being rebuilt. A structure that must be redrawn every time the market moves has the market conditions of one particular year written into it.

Interfaces, not boxes

Executives reach for the chart because it is the only artefact of structure they can actually edit. But most coordination cost does not live in reporting lines. It lives in the interfaces between teams: what one team can promise another, how that promise is versioned, and what happens when it changes. Loose coupling simply means a team can change how it works internally without requiring four other teams to change on the same date.

The idea began in engineering and is usually dismissed as an engineering concern, which is a mistake. It applies to finance, legal and commercial functions with equal force. The useful question to ask of any structure is not who reports to whom. It is how many teams must move together for a given change to happen. When the answer is two, the chart can stay as it is for years. When the answer is eleven, no chart will save you.

A stable interface between teams outranks a tidy reporting line
Count the teams that must move together — that is the real structure
Version what teams promise each other instead of renegotiating it

A structure that has to be redrawn every time the market moves has the conditions of a past year written into it.

NORD/ONE, Organisation practice

Share of structural changes absorbed inside existing team interfaces rather than through a central reorganisation, across seven quarters.

When change management is a confession

Change management as a discrete programme — the workstream, the communications cascade, the readiness survey — is treated as professional practice. It is worth reading more literally. A change that requires a translation layer between the people who designed it and the people who will live inside it was designed without them. The programme is not the cure. It is the invoice for an earlier decision.

The alternative is not less communication. It is earlier participation and a smaller unit of change. Where the affected teams help specify the new interface — what they will owe each other, and by when — the announcement becomes a summary of decisions people already recognise. Our position is blunt: most of what is budgeted for change management would do more good spent on making each change small enough that it does not need managing.

Fatigue comes from surprise, not frequency

Change fatigue is real, and it is routinely misdiagnosed as too much change. In the organisations we have measured, the correlation with frequency is weak while the correlation with unpredictability is strong. People absorb a great deal when they can see the shape of what is coming and, more importantly, when they are told what will not move. What exhausts them is learning on a Tuesday that their reporting line changed on Monday.

Nor is adaptation a smooth ascent. The seven quarters below dip twice: once early, while the first interfaces were still being negotiated, and again in the fifth quarter, when an acquisition brought in teams with entirely different assumptions about what a promise between functions obliges. Both recovered. The trend is what matters, and organisations that panic at a single quarter of decline tend to reach for the one instrument they know how to use.

NORD/ONE is a fictional company. Figures in these articles are illustrative.

Thinking about a move like this?

We work with leadership teams on exactly these questions.

Let's talk

Continue reading

Global Business