Expertise · Change management
Defending a direction is one job. Organising it is another.
Most change programmes spend their energy explaining why the change is right. The people who must carry it out are asking something else: what happens to my work, my standing and my colleagues? “Resistance” is the label put on the gap between those two conversations.
Not every objection is resistance
People may object because they do not understand a change, but also because they see a problem with its implementation that others have missed.
Illustration · where a change stalls
From a decision to a way of working
Two permit teams, one new case system, and a rollout that has stalled
A fictional municipal permit department merges two teams under new case-handling software. Three months in, the system is installed and staff still keep parallel spreadsheets. The project lead calls it resistance. The sponsor wants a communication campaign and mandatory training. Meanwhile, the most trusted caseworker from the older team says openly that the system routes complex dossiers to the wrong people.
Is she resisting, or is she right? The two labels lead to opposite actions, and the word “resistance” makes it unlikely that anyone checks.
Why the rollout has stalled
They can coexist. The first and fourth are about the plan, the second and third about how people are treated. Only if all four are ruled out does the word “resistance” add anything.
1 · The change is partly wrong
Mechanism: the dissent carries information about a real defect.Cue: objections are specific, repeatable and checkable (misrouted dossiers can be counted), come from experienced people, and would stop if the defect were fixed.Move: test the claim against the data, fix or explain, and credit the person who raised it.2 · The reasons are not visible
Mechanism: the reasons are known to those who decided and to nobody else.Cue: ask five people why the change is happening. You get five versions, or “because management says so”.Move: let the affected people test the reasons, not only hear them, and publish one page on what is fixed and what is open.3 · Loss of control, standing or networks
Mechanism: the change moves who decides, who counts as senior and who sits with whom, and the loss is felt unevenly.Cue: objections concern decision rights and seniority more than the software, and informal leaders are quiet or lukewarm.Move: map who loses what, keep what can be kept, give real choice where there is any, and name the loss out loud.4 · Implementation design
Mechanism: the sequence, the workload or the tools make the new way harder than the old one.Cue: use varies by team and by week, not by attitude. Training came before the system was usable. Parallel work doubles the load.Move: change the sequence, pilot in one team, and cut the double work.Two jobs that are easily confused
Defending the direction
The sponsor’s job: the reasons, the mandate, consistency over time. It answers “why this?”
Organising the delivery
Decision rights during the rollout, a pilot, indicators people can influence, routes for feedback, and agreed adjustments. It answers “how, and what do we do when it does not work?”
Confusing those tasks can leave delivery problems unresolved. The sponsor explains the direction for the fifth time while the delivery problems go unaddressed.
The work
What we need to test during implementation
- Separate what is fixed from what is openWe write down what has been decided and cannot move, and where there is real room. “Decided, with real room to shape the how” is honest. A consultation on a finished plan is not. You get: a one-page statement of fixed and open.
- Map the objections and the networksWe collect objections one by one and sort them into factual claims about the plan, requests for information, losses and workload problems. We also find out whom people actually listen to. You get: sorted objections and a map of informal influence.
- Test what can be testedWhere an objection is a factual claim, we look at the data. You get: a list of objections that were right, wrong or untestable.
- Design a pilot with process indicatorsOne team, one process, reversible. Indicators the team can influence, agreed with them, and never turned into a personal scoreboard. You get: a pilot design with predictions and stop rules.
- Run the loopFeedback at the rhythm of the work cycle, adjustments written down, every decision communicated with its reason. You get: a short log from the pilot to the wider rollout.
Pilot rules, agreed before it starts
Decision rules
- ContinueThe process indicators reach the agreed level and none of the named side effects appears.
- AdjustThe indicators fall short, but the objections point to a cause that can be fixed within the pilot.
- ReverseA defect named by the team is confirmed and cannot be fixed in the window.
- Keep two questions apartWhether the decision was right is judged on outcomes. Whether it is being carried out well is judged on process. Mixing them lets a good decision be blamed for poor delivery, or the other way round.
- FairnessNobody is measured on something they cannot influence.
You will often read that most change programmes fail, usually with a precise percentage. We do not use those figures. They are repeated far more often than they are sourced, and “failure” is defined differently each time.
How certain are we?
What research tells us about change
- EstablishedParticipation in planning, perceived fairness and the experience of having some control are associated with greater acceptance of change.
Associations are well replicated. How much comes from participation itself is harder to say, and token participation does not produce the effect.
- Plausible applicationSmall reversible pilots with process indicators that teams can influence find defects earlier than a single large rollout.
Standard improvement practice with mixed evidence on size of benefit. We test it in your setting.
- Our hypothesisSorting objections into four kinds before responding shortens the time to the right move.
Limits of change advice
- A universal phase model of change. Such models describe, and they have not been shown to predict who will do what.
- A communication campaign as a first move. If the defect is real, louder explanation makes things worse.
- A way of predicting who will resist from personality. We look at situations, interests and losses.
- A judgement on whether the change is a good idea. When that is the open question, it calls for a Decision Review first.