Do Change Management Models Actually Work?

The "70% failure" myth is perhaps the most consequential unsubstantiated claim in management practice. The statistic is routinely cited in consulting pitches, textbooks, and boardrooms, yet its origins are remarkably shaky. Researchers tracing the claim have found it loops back to a 1993 book by Michael Hammer and James Champy on reengineering, which offered the figure without empirical evidence. It was subsequently repeated by John Kotter, McKinsey, and others until it became received wisdom. Hughes (2011) conducted a detailed genealogy of the claim and concluded that there is "no valid and reliable empirical evidence" to support a 70% failure rate for organisational change. The figure persists because it serves the commercial interests of change consultancies and because failure makes a more compelling narrative than qualified success.

Jones, Firth, Hannibal and Ogunseyin (2019) published the most comprehensive analysis to date, examining approximately 200 case studies of organisational change. Their findings directly challenged the failure narrative. When outcomes were assessed against the specific objectives of each change initiative — rather than against an idealised standard of total transformation — most changes achieved meaningful progress. The authors found that binary framing (success versus failure) obscures the reality that change outcomes exist on a spectrum. Partial successes, delayed benefits, and unintended positive consequences are common but are routinely miscategorised as failures.

Lewin's Three-Step Model (1951) — unfreeze, change, refreeze — is the foundational framework taught in virtually every organisational behaviour course. Its elegance is also its limitation. Lewin's original work was rooted in small-group dynamics and social psychology experiments, not large-scale organisational transformation. The model provides a useful conceptual metaphor but was never empirically tested as a complete implementation methodology. Burnes (2004) argued that Lewin's work has been oversimplified by subsequent authors — Lewin himself emphasised the complexity and iterative nature of change, but the three-step reduction strips this nuance away. Critically, the "refreeze" concept has been challenged as unrealistic in environments where continuous change is the norm.

Kotter's 8-Step Model (1996) — from creating urgency through to anchoring changes in culture — is arguably the most widely adopted change framework in practice. It emerged from Kotter's observations of over 100 organisations, published first as a Harvard Business Review article and then as a bestselling book. However, the model was developed inductively from consulting experience, not from controlled research. Appelbaum, Habashy, Malo and Shafiq (2012) conducted a systematic review of the evidence for each of Kotter's eight steps individually. They found theoretical and some empirical support for individual steps — for example, the importance of establishing a sense of urgency and building a guiding coalition — but no evidence that following all eight steps in sequence produces better outcomes than alternative approaches. The model has not been tested as a complete system in experimental or quasi-experimental designs.

ADKAR (Awareness, Desire, Knowledge, Ability, Reinforcement), developed by Prosci, is the most commercially successful individual change model. It focuses on the individual level — each person must move through five stages for change to succeed. Prosci's own benchmarking data, drawn from thousands of participating organisations, suggests that projects with excellent change management are six times more likely to meet objectives. However, this data comes from self-selecting participants in Prosci's research programme and is not published in peer-reviewed journals. Independent academic validation of ADKAR as a complete model is sparse. Among the models reviewed in comparative analyses, ADKAR tends to rate highest on adaptability — its individual-level focus makes it easier to apply across different types of change — but this flexibility has not been shown to translate into superior outcomes.

Emergent and complexity-based approaches have gained theoretical traction as alternatives to prescribed sequential models. These draw on complexity theory, systems thinking, and sensemaking to argue that organisational change is inherently nonlinear and context-dependent, making step-by-step recipes inappropriate. Scholars like Stacey, Snowden, and Weick have provided compelling theoretical frameworks, but large-scale empirical testing of complexity-based change approaches remains limited. The approach is intellectually persuasive but practically underdeveloped.

The overarching finding from the research is stark: no single prescribed sequential model has demonstrated superiority over alternatives in experimental or quasi-experimental research. The models that dominate practice were built from observation and consulting wisdom, not from the kind of controlled testing that would establish causal efficacy.

Stop citing the 70% failure rate. It is not supported by credible evidence and creates a misleading baseline. Using it in business cases or board presentations undermines your credibility with anyone who checks the source. If you need to make the case for investing in change management, use the genuine evidence — that thoughtful, well-resourced approaches to change produce better outcomes than unmanaged change — without inflating the risk.

Use models as frameworks, not recipes. Kotter's steps, Lewin's phases, and ADKAR's individual focus all contain useful conceptual insights. Creating urgency matters. Building coalitions matters. Attending to individual readiness matters. But treating any of these as a rigid sequential checklist to be followed mechanically is not supported by the evidence. The research suggests that context — industry, organisational history, type of change, workforce characteristics — determines which elements matter most.

Evaluate change on a spectrum, not in binary terms. The Jones et al. analysis demonstrates that most change initiatives achieve partial success. Define success criteria before launching the initiative, measure progress against those specific criteria, and resist the temptation to declare failure because the outcome was not identical to the original vision. Adaptation during implementation is not failure — it is how complex change actually works.

Invest in the fundamentals that do have evidence. While no complete model has been validated, individual elements consistently appear in successful change efforts across the literature: senior leadership commitment, clear communication, employee participation in design, adequate resourcing, and attending to individual readiness. These are the evidence-based "active ingredients" regardless of which branded model you use.