The dominant change management frameworks — Kotter's 8 Steps, ADKAR, Prosci — treat change primarily as a communication and buy-in challenge. If you communicate effectively, involve people early, create a coalition of support, address resistance, and celebrate wins, change will stick. The logic is compelling. It maps onto what we know about human psychology and organizational culture. It also consistently underperforms.
The gap between what these frameworks promise and what organizations experience is not a coincidence. It reflects a structural omission: the frameworks are largely silent on governance. Who has the authority to mandate the change? What happens to people who do not comply? How is the changed state maintained once the change management effort ends and the consultants go home?
These are not supplementary questions. They are the questions on which change lives or dies. Communication without authority is advocacy. Buy-in without consequences is preference. Adoption without maintenance is a temporary behavioral change that reverts to equilibrium when the pressure lifts.
I have observed change programs that were executed exactly as the frameworks specify — full stakeholder communication, involvement at all levels, trained change champions, visible executive sponsorship — and that still produced organizational regression within six months of the formal program's conclusion. In most of those cases, the explanation was the same: the governance architecture that would have sustained the change was never built.
Why Change Management Fails Despite Good Communication
The communication-centric change management tradition produces recognizable failure modes. Each one is a governance failure disguised as a communication failure.
Authority is assumed, not established. Most change programs proceed as though the authority to mandate the change exists by virtue of the organizational position of the people sponsoring it. Senior leaders announce the change; it is assumed that the announcement carries authority. This is frequently not true in practice. In organizations with strong informal authority structures, longstanding norms, or professional cultures that have their own sources of legitimacy, a senior leader's announcement is the beginning of a negotiation, not the establishment of a mandate. The change program then runs as a communication effort when what it actually requires is a renegotiation of authority — a substantively different kind of work.
Consequences are not enforced. Most change programs that specify consequences for non-compliance do not enforce them consistently. The reasons are understandable: enforcement is uncomfortable, the relationship costs feel high, and in the early stages of a change program there is usually a preference for positive messaging. But consequence-free non-compliance sends a clear signal to everyone watching: the change is optional in practice regardless of what the formal communication says. Once this signal is established, it is very difficult to reverse.
Old systems remain available as fallbacks. Change programs that do not remove access to the old way of doing things will see reversion. If the new process is more complex than the old one, and the old one is still available, people under pressure will use the old one. The change management framework typically treats this as a resistance and communication problem — find out why people prefer the old system, address their concerns. It is actually a governance problem: the governance architecture has not made the old system unavailable. Making the changed state the only available option is often more powerful than any amount of additional communication.
Incentive structures are not updated. People are evaluated, compensated, and promoted based on metrics that existed before the change. If the change requires different behavior, and the evaluation system still rewards the old behavior, the evaluation system wins. This is not a communication problem. It is a structural problem. The governance architecture — the system of incentives that shapes day-to-day behavior — has not been aligned with the change that the communication campaign is trying to produce.
Change management ends before change is sustained. Change programs have budgets, timelines, and conclusion events. The program ends; people move on to other priorities; the change management infrastructure is dismantled. Whether the change has actually been embedded into the operating system of the organization — into its processes, its governance, its incentive structures — is often not evaluated at program conclusion. Six months later, when the behavioral regression is visible, the explanation is attributed to culture or to the difficulty of the change, not to the fact that the governance infrastructure required to sustain it was never built.
How a Well-Communicated Change Reverts: A Worked Example
It is worth tracing a single reversion in detail, because the mechanism is more instructive than the summary. Consider a common change: an organization moves from an informal, email-based approval process to a structured request system that creates an audit trail. The communication is exemplary — leadership explains the regulatory exposure the old process created, champions are trained in each department, a launch event marks the cutover, and first-month adoption is strong.
Then watch the governance seams. In week three, a manager under deadline pressure cannot get a request approved fast enough, so she emails the approver directly; the approver, who values the relationship, says yes, and no consequence follows because no one specified one — the first gap. Because the email path worked once, it works again, and by week eight it is the fast lane while the new system is the slow lane. The old system was never decommissioned — the second gap — so it stayed available, and availability under pressure is destiny. Managers are still evaluated on cycle time, not audit-trail completeness — the third gap — so the old behavior keeps the structural support the new behavior never received. By month six the program team has moved on, no one owns sustained compliance — the fourth gap — and a review finds a third of approvals never entered the new system. The post-program narrative blames "change fatigue." It was not fatigue. It was four unmade governance decisions, each surfacing at a predictable seam, none of which the communication campaign was designed to address. Communication produced understanding and even enthusiasm; none of it mattered at the seams, because the seams are governed by authority, consequences, availability, and incentives — not by understanding.
The Governance Decisions That Have to Be Made Before Change Management Begins
There is a set of governance decisions that must be made before a change program begins — not during the communication phase, but before the first meeting with stakeholders about what is changing and why.
Who has the authority to mandate this change? This is the foundational question. The answer determines whether the change program is an advocacy effort (here is why you should adopt this change) or an implementation effort (this change is being made; here is how). The two programs look superficially similar but are fundamentally different. An advocacy effort can succeed, but it succeeds when the argument is persuasive. An implementation effort succeeds when the implementation is executed. Clarity about which type of program you are running shapes every subsequent decision.
What happens to people who do not comply, and who enforces it? This question feels uncomfortable in the context of change management, where the communication tradition emphasizes inclusion and buy-in. But it is the question that separates changes that stick from changes that revert. The consequence chain must be specified before the program begins, communicated clearly during the program, and enforced consistently after it. If the organization is not prepared to enforce the consequence chain, the change should be redesigned as an advocacy effort, not a mandatory implementation.
Which old systems will be decommissioned, and when? If the change requires new systems or processes, and the old ones remain available, the change will not fully take hold. The plan for decommissioning old systems — with clear timelines and without exceptions for people who are not yet comfortable with the new — is a governance decision that must be made before the program begins. Leaving it to be decided later almost always results in the old systems remaining available indefinitely.
Which incentive structures will be updated to reflect the changed state? Identify the specific evaluation criteria, compensation components, and cultural rewards that currently support the old behavior and that will need to be updated to support the new one. Create a plan and timeline for those updates. This is governance work that is typically outside the scope of a change management program budget, which is one of the reasons it does not get done.
Who owns sustained compliance after the program ends? Specify the role — the function or the person — responsible for monitoring and maintaining compliance with the changed state after the formal change program concludes. Without a clear owner, sustained compliance defaults to no one, and the organizational pressure to revert will eventually win.
How to Design a Change Program with Both Communication and Governance Backbone
A change program that is designed to produce durable change integrates the communication elements of the traditional frameworks with the governance backbone that they omit.
The communication elements remain important. People who understand why a change is being made are more likely to implement it well, to identify problems early, and to adapt as the implementation encounters reality. Involving people in the design of implementation details — not in the decision of whether to change, but in how to change — produces better implementation quality and stronger ownership. Visible executive sponsorship signals organizational priority in a way that middle-management communication cannot replicate.
The governance backbone provides the structural conditions under which the communication is sufficient to produce change rather than producing only understanding.
Stage 1: Governance architecture before communication begins. Before the first stakeholder meeting, the five governance decisions above should be resolved. The authority mandate should be established. The consequence chain should be specified and reviewed by the people who will enforce it. The decommissioning plan for old systems should be approved. The incentive structure updates should be scoped and scheduled. The sustained compliance owner should be identified.
This stage takes longer than most organizations expect. It requires decisions that are politically difficult — about who has authority, about what the consequences of non-compliance will be, about whose evaluation criteria will change. Organizations that skip this stage and go directly to communication are committing to a change program that is structurally dependent on the persuasiveness of its communication. That is a fragile foundation.
Stage 2: Communication designed around the governance architecture. With the governance architecture in place, the communication is designed to be honest about it. People are told not just why the change is being made, but who has the authority to mandate it, what happens if they do not comply, when the old systems will be decommissioned, how their evaluation criteria will change, and who is responsible for sustained compliance. Honest communication about the governance architecture is more effective than communication that avoids it, because people know the governance architecture will shape their experience regardless of whether it is communicated.
Stage 3: Implementation with governance enforcement. As the change is implemented, the consequence chain is enforced consistently from the beginning — not after a grace period, not after repeated warnings that are not followed by consequences. Consistent early enforcement establishes the changed state as real rather than as aspirational. Inconsistent early enforcement establishes the changed state as optional, which is extremely difficult to reverse.
The implementation phase also monitors for the specific failure modes that the governance architecture was designed to prevent: people reverting to old systems, metrics being gamed to appear compliant without actual behavioral change, informal workarounds that reproduce the old behavior in new forms. These are not communication problems at this stage; they are governance problems. They require the enforcement mechanisms that the governance architecture specified.
Stage 4: Transition to sustained compliance. The formal change program ends not when the communication campaign concludes, but when the changed state is embedded into the operating system of the organization — into its processes, its governance, its incentive structures, its monitoring and reporting. The transition to sustained compliance is an explicit event, with explicit handoff to the sustained compliance owner, with explicit review of whether the governance architecture is in place and functioning.
What Sustained Change Requires Structurally
The question of what makes change durable — as opposed to what makes it initially adopted — is rarely asked in the change management literature. Initial adoption and sustained change are different problems, and the tools that produce initial adoption do not necessarily produce sustained change.
Sustained change requires four structural conditions.
The changed behavior must be the path of least resistance. People under pressure default to whatever is easiest. If the new process is harder than the old one — even slightly — people under pressure will find ways to use the old approach. Sustained change requires that the changed behavior be made easier than any alternative. This means removing access to alternatives, redesigning the new process to eliminate unnecessary complexity, and investing in the tooling and support that makes the new behavior genuinely easier.
The incentive structure must support the changed behavior. As long as the evaluation system rewards the old behavior, the old behavior has structural support that the new behavior does not. Sustained change requires that the incentive structure be updated before the change is considered complete — not scheduled for eventual update, but actually updated.
The monitoring and feedback system must detect and address regression. Organizations that are not actively monitoring for regression will not notice it until it is advanced. Sustained change requires a monitoring system that detects early signs of reversion — through process data, through exception reporting, through regular compliance review — and a feedback loop that surfaces regression to the sustained compliance owner quickly enough to be addressed before it becomes entrenched.
The organizational story about the change must be accurate. Organizations develop narratives about the changes they have made. When those narratives attribute difficulty to external factors rather than to the governance gaps that actually caused them, the organization does not learn the lessons that would make the next change more successful. Accurate retrospective accounting of what worked and what did not — including honest acknowledgment of where the governance backbone was absent — is a structural condition for organizational learning.
Where Governance-Heavy Change Becomes the Wrong Tool
The governance backbone is not free, and presenting it as always-correct would be its own kind of dishonesty. There are contexts where building the full architecture costs more than it returns.
The clearest is the genuinely reversible, low-stakes change. If the cost of reversion is small and the change can be re-attempted cheaply, the overhead of establishing authority mandates, consequence chains, and decommissioning schedules may exceed the cost of simply trying the change and adjusting. Governance rigor is proportional to the cost of failure; applying maximum rigor to a trivial change is the bureaucratic mirror image of applying no governance to a critical one. The same caution applies to early, exploratory change where the right end state is not yet known — decommissioning old systems and hard-wiring incentives presume you already know what the new state should be, and unwinding a fully governed change is more expensive than unwinding an informal one. In exploratory phases, keeping optionality open is the correct call; governance comes after the destination is settled, not before.
The honest position is that governance is a cost incurred to buy durability under conditions where communication cannot reach — distance, scale, and a large number of independent actors. In a small organization where one person can mandate the change, observe compliance directly, and adjust incentives by conversation, proximity already provides what the formal architecture would. When durability is cheap or those conditions are absent, the cost is not justified. The skill is matching the rigor to the stakes, not maximizing rigor everywhere.
The Leadership Implication
The implication for leaders is uncomfortable: change management, understood correctly, requires leaders to make governance commitments that many of them would prefer to leave ambiguous.
It requires leaders to establish clear authority — which means being explicit about who can mandate the change and who cannot, which surfaces the political dynamics of authority that leaders often prefer to manage informally. It requires leaders to specify and enforce consequences — which means being prepared to follow through on consequence chains that may affect people with whom they have relationships. It requires leaders to decommission old systems on schedule — which means resisting the pressure from people who are not yet comfortable with the new.
None of this is comfortable. All of it is necessary for change that is meant to be durable.
The leaders who consistently produce sustained organizational change — not the leaders who produce impressive initial adoption followed by gradual regression, but the leaders whose organizations actually operate differently a year after a significant change — are the ones who understand that communication creates understanding and governance creates behavior. Both are necessary. Neither alone is sufficient.
Conclusion
Change management has been treated as a communication discipline for long enough that the governance omission has become invisible. The frameworks are mature, the tools are well-developed, and the practitioners are skilled. The outcomes are consistently disappointing.
The path to better outcomes is not better communication. It is the governance architecture that makes communication sufficient: clear authority, specified and enforced consequences, decommissioned alternatives, updated incentive structures, and a sustained compliance owner who holds the changed state after the program ends.
If you have one change in flight right now, the most useful thing you can do this week is not to refine its messaging. It is to write the five governance decisions on a single page and answer them honestly: who can mandate this, what happens to non-compliance and who enforces it, which old system gets decommissioned and when, which incentive changes by what date, and who owns compliance after the program ends. Wherever the page is blank, you have located a seam where the change will revert — before the reversion happens, while it is still cheap to fix.
This is more work than a communication campaign. It is also the work that actually produces change.
Continue in this series
This piece is part of What Is Organizational Governance? A Systems Practitioner's Complete Guide, my systematic guide to organizational governance and operating systems. Related reading:
- Organizational Mapping as a Governance Tool
- Operating Distributed Teams Across Philippine Time and Geography
- How to Anticipate Unintended Consequences Before They Happen
- Knowledge Management as Governance Infrastructure
Working through this in your own organization? I help technical leaders design it directly — advisory engagements.
