Resistance as Data: What the Pushback Is Actually Telling You
Pushback is not the failure of communication; very often it is communication — the last channel a programme leaves open to the people it is being done to.
Executive Summary
Every change programme generates a stream of information it did not commission and does not want: the objection raised for the third time, the meeting that goes quiet, the workaround that reappears a fortnight after it was banned. We are trained to call this resistance, and trained to overcome it. The argument of this essay is that the instinct to overcome is often the instinct to stop listening at precisely the moment the most valuable data arrives. Resistance is not the absence of information about a change. Frequently it is the densest concentration of information available — the accumulated, local, uncodified knowledge of the people closest to the work, expressed in the only register the programme has left open to them.
To treat pushback as a signal to be read, rather than a barrier to be removed, is not softness. It is a more rigorous form of diagnosis. It does not mean deferring to every objection, and it does not mean that all resistance is wisdom in disguise — some of it is fear, some of it is self-interest, some of it is simple fatigue. It means refusing to discard, unexamined, the one dataset a programme cannot buy from an advisory firm and cannot generate from a plan. The discipline this essay describes is not learning to like resistance. It is learning to read it before deciding what to do about it.
The report that reads green
Picture a steering meeting late in the life of a shared-services consolidation. The finance operations of a dozen countries are being pulled into a single standard process; the business case promised a shorter month-end close and a lower cost per transaction, and the programme is, by every measure on the slide, on track. The status is green. The milestones are met. The training completion figures are respectable. And for the fourth month running, the same regional controller raises the same objection about the same reconciliation step, and for the fourth month running it is logged as a local exception, tagged as a training issue, and closed.
Everyone in that room has been taught how to categorise what they are hearing. It is resistance. It is the natural friction of change. It is, the change-management toolkit reassures us, to be expected around the trough of the transition curve, and the correct response is reinforcement: clearer communication, a firmer message from the sponsor, perhaps a quiet word about alignment. The one interpretation almost nobody reaches for is the literal one — that the controller is not resisting the change so much as reporting a fact about the work that the design does not yet accommodate.
This is the quiet tragedy of a great deal of change practice. We have built an elaborate apparatus for managing resistance and almost none for reading it. We map stakeholders by their influence and their disposition, we score them from champion to blocker, we plan our interventions accordingly — and in doing so we convert every human objection into a problem of positioning rather than a possible statement of fact. The blocker, by definition, is someone to be moved. It rarely occurs to us that the blocker might be right.
Two ways to hear a “no”
There are, broadly, two ways to hear an objection, and the whole argument turns on the difference between them.
The first treats the “no” as a state to be changed. On this view, the person objecting has a deficit — of understanding, of buy-in, of will — and the programme’s job is to close that deficit. The vocabulary gives the game away: we speak of overcoming resistance, of winning hearts and minds, of getting people on board, as though the destination were fixed and the only question were how many passengers we could persuade to board the train before it left. Every one of these phrases assumes the plan is sound and the objection is noise.
The second treats the “no” as a message to be decoded. On this view, an objection is a compressed piece of information about the gap between the plan on paper and the work as it is actually performed. It may be a poor message — garbled, self-serving, exaggerated — but it is a message, and the first move is to ask what it is trying to say before deciding whether to act on it. This is not a softer stance. It is, if anything, more demanding, because it refuses the comfort of dismissing opposition and insists instead on the harder work of interpretation.
The manager who asks “how do I overcome this resistance?” has already decided the resistance is wrong. The more useful question is prior to that one: “what does this person know that my plan does not?”
The distinction matters because the two stances are self-confirming. If you believe resistance is a deficit, you will invest in persuasion, you will measure success by compliance, and you will experience every objection as an obstacle — and you will very often win, in the narrow sense that the change goes in and the objectors fall silent. What you will not discover is whether their silence was agreement or surrender, and whether the thing you pushed through was improved or merely imposed. The cost of that stance is invisible precisely because it succeeds.
What the pushback encodes
Pushback is not the failure of communication; very often it is communication — the last channel a programme leaves open to the people it is being done to. Once we start reading it that way, the objections that arrive during a change effort resolve into several distinct kinds of information, each worth something different.
- Knowledge of the work as performed. The people doing a job hold a vast reserve of undocumented knowledge about how it actually gets done — the exceptions, the dependencies, the informal handshakes between teams that no process map records. When a standardised design collides with this knowledge, the collision surfaces as resistance. The objection is the map of what the design left out.
- Knowledge of the last time. In most large organisations, the current change is not the first. People carry the scar tissue of previous programmes that promised much and delivered disruption, and their reluctance often encodes a precise memory of how the last effort failed. What sounds like cynicism is frequently pattern recognition.
- A signal about cost and load. Sometimes the objection is really about capacity — the change is sound, but it is the fourth simultaneous initiative landing on the same exhausted team, and the resistance is the sound of a system at its limit. That is data about sequencing, not about the merit of the change.
- Genuine self-interest. And sometimes the objection is exactly what it looks like: a defence of territory, status, headcount, or comfort. This too is information — about incentives, about who has something to lose, about where the design threatens the existing distribution of power — even if it is not information that should change the plan.
The point is not that every objection belongs to the first category. It is that we cannot know which category an objection belongs to until we have read it, and the standard machinery of resistance management is designed to skip the reading. It sorts objectors by disposition and prescribes an intervention, and in doing so it treats the controller defending a genuine dependency and the manager defending an empire as the same problem, to be solved the same way. They are not the same problem. Telling them apart is the entire skill.
The close that would not shorten
Return to that reconciliation step. In the case I have in mind — a composite, but true to a dozen I have watched — the central design projected the monthly close falling from eleven working days to six, a saving the business case leaned on heavily. One region kept objecting that a particular reconciliation could not simply be dropped. The programme heard resistance: a controller reluctant to let go of a familiar routine, protective of a task that justified a role. It logged the objection as a local exception, offered additional training, and closed the item three times.
What the controller knew, and could not get heard, was specific. That reconciliation was not housekeeping. It fed a statutory filing peculiar to that jurisdiction — a local regulatory adjustment that had to be reconciled before the return could be submitted — and the new standard process, designed around the majority of regions where no such filing existed, had nowhere to put it. The objection was not a preference. It was a fact about the law in one country, arriving in the disguise of an awkward middle manager who would not get on board.
| Measure | What the central design assumed | What actually happened |
|---|---|---|
| Monthly close duration | Eleven days falling to six, everywhere | Seven days in four regions; the fifth could not close cleanly at all |
| The disputed reconciliation | A local habit, removable with training | Fed a statutory filing; its removal made the return late |
| Rework | None budgeted | Two quarters of manual workaround, exceeding the projected saving in that region |
The numbers are modest, and that is the point — this was not a catastrophe that anyone would write a post-mortem about. The close did shorten in most of the estate; the programme could still, honestly, report a win. But in one region a statutory return went in late, a small penalty and a larger conversation with the regulator followed, and the “saving” was consumed for two quarters by the manual patching required to reinstate, off-system, the very step the design had removed. Every cost of that outcome was foreseeable, because it had been foreseen — three times — and filed as resistance.
What is instructive is not that the design was wrong. Designs are always incomplete; that is why we pilot and iterate. What is instructive is that the organisation held the correcting information the entire time, in the form of an objection, and its change apparatus was configured to discard exactly that information. The resistance was the highest-quality data the programme received. It was also the data the programme was least equipped to hear.
The strongest case for pushing through
It would be dishonest to leave the argument there, because there is a serious case on the other side, and any practitioner who has run a large programme has felt its force.
The case is this. Organisations are not short of objections; they are drowning in them. Every significant change threatens someone, and if leaders treated every “no” as a signal to stop and investigate, nothing would ever move. Transformation requires the will to proceed against opposition, because a great deal of opposition is precisely the inertia, the territoriality, and the fear that the change exists to overcome. There is such a thing as a burning platform, and on a burning platform the leader who pauses to fully honour every objection is not being wise; they are being negligent. Sometimes the job is to decide, absorb the resistance, and drive the change through — and the leaders we most admire are often those who did exactly that, over the loud objections of people who turned out to be wrong.
This is not a weak argument, and the reading I am proposing has to answer it rather than wish it away. The answer is that reading resistance and overriding it are not opposites; they are sequential. The failure in the shared-services case was not that the programme eventually decided to standardise despite an objection. The failure was that it never read the objection at all before overriding it. It skipped the diagnosis and went straight to the verdict.
“The choice is never between honouring resistance and having the resolve to act. It is between acting on what the resistance told you and acting in ignorance of it.”
A leader who has genuinely read an objection, understood the knowledge it carries, weighed it, and then decided to proceed anyway has done something entirely different from a leader who dismissed it as noise. The first may still be wrong, but they are wrong with their eyes open, and they will see the consequence coming and can prepare for it. The second is wrong blind. The steelman for pushing through is really an argument for decisiveness, and decisiveness is not in tension with reading resistance. It is in tension only with obeying it — and nobody serious is arguing that we should obey it.
Reading without romanticising
The opposite error is just as real, and worth naming plainly, because the argument of this essay is easily sentimentalised. To say that resistance carries information is not to say that the resisters are always right, that the frontline always knows best, or that consensus is the test of a good design. That would simply invert the old prejudice: instead of assuming the plan is sound and the objection is noise, it would assume the objection is wisdom and the plan is imposition. Both are lazy. Both spare us the work of discrimination.
Reading resistance well means holding two questions apart. The first is diagnostic: what does this objection know? The second is evaluative: given what it knows, what should we do? Confusing the two is what produces both errors. The programme that never asks the first question overrides good information as though it were bad. The leader who collapses the second into the first — who assumes that because an objection contains truth it must therefore be heeded — surrenders the design to whoever shouts. The skill lies in taking every objection seriously as evidence while taking almost none of them as instruction.
There are rough tests that help. An objection that gets more specific under questioning — that resolves into a dependency, a date, a number, a named consequence — is usually carrying knowledge. An objection that gets vaguer, that retreats from the particular to the general the moment you probe it, is usually carrying something else: discomfort, fatigue, or interest. The controller who can tell you exactly which filing breaks and by how many days is giving you data. The manager whose objection dissolves into “the timing isn’t right” and “people aren’t ready” the moment you ask what specifically breaks is telling you about their disposition, which is worth knowing but is a different kind of thing. The register of an objection — whether it thickens or thins under scrutiny — is often a better guide to its value than the seniority or the sentiment of the person raising it.
What this asks of leaders
If resistance is data, then the change leader’s core competence is not persuasion but interpretation, and that is a considerable shift in temperament. We are, as a profession, fluent in method — in the phased plan, the stakeholder map, the communication cascade, the readiness assessment. We are far less fluent in the humbler discipline of treating opposition as testimony and ourselves as its readers rather than its overcomers.
The shift shows up in small, concrete practices. It means designing a programme so that objections have somewhere to go other than the risk log where they are quietly closed — a route by which a frontline “this won’t work” can reach the design authority intact rather than being filtered, at each level, into something more palatable. It means resisting the temptation to measure change success by the speed with which resistance subsides, since the fastest route to quiet is not persuasion but resignation, and a programme optimised for silence is optimised for the wrong thing. It means, above all, treating the moment of strongest objection not as the low point of the transition to be endured but as the richest sampling opportunity the whole effort will offer — the moment when the organisation is telling you, more honestly than any survey will, where your plan meets the world and loses.
None of this is comfortable, because it requires the people who designed the change to accept that the people who object to it may know something they do not, and to build the machinery to find out what. That is a concession of authority that most programmes are structured to avoid. But the concession is not weakness. The organisations that treat resistance as an obstacle to be cleared and the organisations that treat it as intelligence to be gathered will, over enough changes, diverge sharply — not because one is kinder, but because one is systematically better informed. The pushback was always going to arrive. The only question is whether we spend our energy suppressing the signal or reading it.
The controller was, in the end, right about the filing. She was also, for a while, filed under resistance. The distance between those two facts is the distance this essay has tried to close.