Change management models compared: Lewin, Kotter’s 8 steps and ADKAR

A new system, a restructure, a different way of selling: most change in organisations depends on people doing something differently from how they did it last month. The CIPD’s change management factsheet notes that change initiatives are often unsuccessful, and that poorly managed change can hurt performance and engagement and lead to unmanageable job stress.

Change models give you a structure for thinking about what people need at each point. Three come up more than any others: Lewin’s three stages, Kotter’s 8 steps and Prosci’s ADKAR. This guide explains each one, compares them side by side, and shows how to choose or combine them.

Lewin: unfreeze, change, refreeze

The model attributed to the psychologist Kurt Lewin, who died in 1947, describes change in three stages. Unfreezing loosens the current way of doing things and the mindset behind it. Changing is the transition, often a period of confusion. Refreezing makes the new way stable, so people do not drift back.

The model is simple, which explains both its popularity and its main weakness. It treats change as a one-off move between two stable states, while many organisations now change continuously.

Its origin is also less tidy than textbooks suggest. Researchers Stephen Cummings, Todd Bridgman and Kenneth Brown found that the three-step model rests on a brief summary in one of Lewin’s last articles, published in 1947, and that others built it out after his death. Their paper appeared in Human Relations. As Victoria University of Wellington reported, Cummings said the misreading “blinkers us from seeing his other, more substantive insights”. The Wikipedia article on Lewin records the same debate.

One of those insights is force field analysis: listing the forces that push towards a goal and the forces that hold it back. It remains a quick way to see why a change is stuck.

Kotter: eight steps for leading change

John Kotter, a Harvard Business School professor, set out his approach in a 1995 Harvard Business Review article, “Leading Change: Why Transformation Efforts Fail“, based on watching more than 100 companies attempt major change. He found that change runs through phases that take time, that skipping steps creates only an “illusion of speed”, and that a serious mistake in any phase can undo earlier progress.

Kotter Inc lists the eight steps as:

1. Create a sense of urgency.
2. Build a guiding coalition.
3. Form a strategic vision.
4. Enlist a volunteer army.
5. Enable action by removing barriers.
6. Generate short-term wins.
7. Sustain acceleration.
8. Institute change.

Kotter Inc says the original linear model came from the book Leading Change and later evolved into 8 Accelerators, described in Accelerate (2014) and Change (2021). The newer framing treats the steps as things that run in parallel, driven by a network of volunteers alongside the normal hierarchy.

Kotter is strongest at the organisational level: building support among leaders, making the case, and keeping momentum. It says less about what each individual needs in order to change how they work.

ADKAR: change one person at a time

Prosci describes ADKAR as a model for individual change. The letters stand for five elements, in Prosci’s words:

• Awareness: understand why the change is needed.
• Desire: choose to participate in and support the change.
• Knowledge: know how to change and what success requires.
• Ability: put new skills and behaviours into practice.
• Reinforcement: sustain the new way and prevent backsliding.

The practical idea is the barrier point. Prosci advises finding the earliest element a person or group has not yet reached and focusing support there. Training (knowledge) will not help someone who does not see why the change is happening (awareness). Prosci notes that the elements are sequential but not strictly linear, and that ADKAR sits inside its wider methodology, which adds sponsorship and communication at organisation level.

The three models compared

Choosing and combining them

The models answer different questions, so many teams use more than one. Kotter helps leaders plan the programme. ADKAR helps managers support their own teams. Lewin gives everyone a shared, simple language, and force field analysis helps when progress stalls.

1. Describe the change in one sentence and list who has to do something differently.
2. Use Kotter’s early steps to check you have urgency, a coalition and a clear vision before launch.
3. For each affected group, assess ADKAR and find the barrier point.
4. Plan support for each barrier: communication for awareness, involvement for desire, training for knowledge, practice and coaching for ability, recognition and measures for reinforcement.
5. Plan short-term wins and share them.
6. Decide how you will know the new way has stuck, and keep measuring after go-live.

An example (illustration only)

A 40-person recruitment agency moves from spreadsheets to a shared CRM.

The managing director uses Kotter’s first steps. She explains the urgency (lost candidates and duplicated calls), forms a small group of respected consultants to lead the move, and sets out a simple vision: one record for every client and candidate. Early wins are shared at the Monday meeting.

Six weeks in, one team still uses spreadsheets. An ADKAR check shows they understand the reason and want the benefits, but lack knowledge: they were on holiday during training. The barrier is knowledge, not desire, so the fix is a short training session and a buddy, not another speech. The agency then reinforces the change by running pipeline reviews only from the CRM.

Common mistakes

• Treating Kotter’s steps as a one-time checklist rather than ongoing work.
• Quoting precise failure rates for change programmes. Kotter’s 1995 article gives no percentage; it says most efforts fell somewhere between success and failure. Prefer evidence from your own organisation.
• Jumping to training before people understand why the change matters.
• Declaring victory at go-live. Reinforcement and refreezing happen after launch.
• Using ADKAR alone for a large programme without sponsorship and communication planning.
• Applying one model rigidly. Each is a thinking aid, not a guarantee.