The Three Things That Make Change Actually Stick

Companies with mature change-management functions, sophisticated communication plans, and well-staffed transformation offices fail in the rollout at roughly the same rate as companies with none of those things. That should be unsettling. It tells us most of what is sold as 'change management' is theatre — useful theatre, sometimes, but theatre.

The three patterns below are not theatre. They are unglamorous, and they predict whether the change held 18 months later better than any maturity model I have seen.

Sustained change has a single human owner who is on the work daily — not the executive sponsor, not the steering committee chair, but the person who would have to explain personally why it failed. In healthy programs, that person is a senior line manager whose day job moves with the change.

When the owner is a transformation-office director with no operating P&L, the change reverts. Operators correctly read the org chart and conclude that the new way is optional. Within 90 days of the launch event, it is.

The second pattern is about ownership of the operating rhythm. In the change programs that hold, the weekly cadence — the standup, the dashboard review, the exception meeting — was designed by the operators themselves, not handed down by the consultants. It is sometimes uglier than what we would have designed. It always survives longer.

This is hard for consulting teams to accept. We are paid to bring rigor. But the cadence is the part of the program that has to live in the organization after we leave, which means the operators have to feel ownership of it. The compromise is to give them a starting template and step out of the way.

The third pattern is the most counterintuitive. Change programs that stick are designed with a visible way to roll back. The operators can name the conditions under which the new approach would be paused, the named decision-maker who could pause it, and what the rollback would actually look like.

This sounds like a way to invite failure. In practice it does the opposite. The fear of being unable to roll back is what kills early adoption — operators sandbag the new way because they cannot tell whether it will work and they do not want to be trapped if it does not. A visible exit removes that fear and, paradoxically, makes the rollback almost never necessary.

When you walk back into a company 18 months after a change program that held, the change is invisible. Nobody talks about it. It is just how the work gets done. The Monday cadence is on the calendar. The dashboard is in the standing slot. The new role descriptions are unremarkable.

Programs that did not stick are also easy to spot. The dashboard is still there but nobody looks at it. The new role exists on paper but the work is being done the old way under a different label. The transformation office is still funded, but the operators have quietly routed around it.

The difference between those two outcomes is rarely the deck or the diagnostic. It is whether the three patterns above were present, six months in, when nobody was watching anymore.