What change management actually addresses

Change management is the coordinated work of altering structures, routines, technologies, roles, or behavior while maintaining the conditions people need to implement the change. It is not synonymous with sending communications, following a fixed sequence, or persuading reluctant employees. The relevant outcome is not whether a launch occurred; it is whether people can perform the new work, whether the system supports it, and whether the intended results persist.

Readiness is especially important. Weiner (2009) defines organizational readiness as a shared psychological state involving collective commitment to implement a change and collective belief in the ability to do so. This makes readiness different from individual enthusiasm (Weiner, 2009). A team can believe that a change is valuable yet doubt that time, skills, authority, or coordination are sufficient. Conversely, people may feel capable of implementing a change they do not consider worthwhile.

Armenakis, Harris, and Mossholder (1993) describe readiness as beliefs and attitudes concerning the need for change and the organization’s capacity to accomplish it. Their work supports a practical distinction (Armenakis et al., 1993): leaders must establish both a credible discrepancy between current and desired conditions and confidence that the proposed response is appropriate and feasible.

What research supports—and what it does not

Employee reactions are not one-dimensional

A major review by Oreg, Vakola, and Armenakis (2011) organizes reactions to organizational change around antecedents, explicit reactions, and consequences. The review shows why labeling all disagreement as “resistance” is analytically weak. Reactions may reflect personality, change process, perceived harm, trust, participation, organizational context, or the content of the change itself. Some objections reveal implementation risks that sponsors have not recognized.

Rafferty, Jimmieson, and Armenakis (2013) integrate readiness research across individual, work-group, and organizational levels. Their multilevel perspective cautions against averaging away important differences (Rafferty et al., 2013). Senior leaders may report high readiness while a frontline unit lacks equipment or a middle-management group faces conflicting incentives.

Participation can improve implementation conditions

Participation is not a ceremonial listening session. It should give people a real opportunity to identify workflow consequences, test assumptions, and influence implementation choices. Stouten, Rousseau, and De Cremer (2018) synthesize change research into evidence-based steps including diagnosing the problem, creating urgency, developing the change vision, building goals, seeking support, planning implementation, and institutionalizing the change. Their review also stresses context: no single recipe guarantees success.

None of this justifies a claim that a particular percentage of changes fail. Definitions of “change,” “failure,” and time horizon vary too widely. A defensible evaluation names the intended outcomes, comparison point, observation period, and adverse effects.

A defensible change process

1. Specify the change and outcome. Describe which decisions, behaviors, systems, and results must differ. Separate outputs such as training completion from outcomes such as error reduction or adoption.
2. Diagnose the current system. Map incentives, workflows, dependencies, informal practices, workload, and prior attempts. Record evidence rather than assuming the problem is attitude.
3. Assess readiness by group. Examine perceived value, efficacy, leadership support, resources, timing, and local constraints. Do not treat an enterprise average as evidence that every unit is ready.
4. Co-design implementation details. Invite affected people to test procedures and identify failure modes. State which decisions are open to influence and which are constrained.
5. Build capability. Provide practice, job aids, coaching, access, time, and technical support. Knowledge without opportunity and system support is not implementation capacity.
6. Pilot and compare. Use a limited deployment when feasible. Establish baseline measures and predefined thresholds for expansion, revision, or rollback.
7. Reinforce and learn. Align supervision, incentives, staffing, and performance systems with the new work. Continue measuring after launch to detect workarounds and unintended effects.

Measurement and decision record

A balanced scorecard for change should combine implementation, experience, and operational evidence (Weiner, 2009). Possible measures include eligible-user adoption, correct-use rate, time to proficiency, exception volume, rework, service outcomes, workload, safety events, and distribution of effects across groups. Survey measures can assess readiness or attitudes, but they should not substitute for behavioral and operational outcomes.

For every measure, record the definition, data owner, baseline period, target or decision threshold, review date, and known limitations. If adoption rises while errors or workload worsen, the implementation is not automatically successful. If a pilot succeeds under unusually high support, leaders should test whether that support can be sustained at scale.

The decision record should also distinguish evidence from judgment. Evidence may show that a unit lacks protected training time; management judgment determines whether to delay, reduce scope, or add resources. Making that distinction visible improves later learning.

Common failure modes

Readiness instruments should be selected for their construct and context rather than treated as interchangeable employee sentiment measures (Holt, 2007). Reviews of change recipients also support analyzing the process, context, content, and individual differences behind observed reactions (Oreg, 2011).

• Using communication volume as proof of understanding or commitment.
• Calling reasonable operational objections resistance.
• Training people before the process, permissions, or technology are stable.
• Measuring launch activity but not correct use, outcomes, or unintended harm.
• Assuming the sponsor’s readiness represents frontline readiness.
• Declaring victory before incentives and routine management practices support the change.

Ethical change practice also considers who bears the cost. A project can improve aggregate efficiency while increasing surveillance, workload, exclusion, or risk for a subgroup. Leaders should provide accessible feedback and appeal channels and should examine distributional outcomes rather than relying only on enterprise averages.

What competent performance looks like

• Defines the desired behavioral and operational change precisely.
• Uses readiness evidence at the level where implementation occurs.
• Treats participation as input to design and risk control.
• Links training to practice, resources, authority, and reinforcement.
• Uses pilots, thresholds, and rollback criteria where feasible.
• Reports outcomes, unintended effects, assumptions, and uncertainty.