The change management process

In plain terms

A change management process is the plan for the people side of a change: how the people affected will learn about it, be prepared for it, get help during it and keep doing the new thing once the attention moves on.

Every named framework has its own vocabulary. Underneath, several well-known ones can be read as following a similar arc, from preparing for the change, through launching it and supporting people, to checking that it has held.

Where it came from

Lewin's three steps. The sequence unfreeze, change, refreeze is usually credited to Kurt Lewin, though Cummings, Bridgman and Brown (2016) argue that Lewin never developed it as a model and that it took shape after his death. See Kurt Lewin.

Kotter's eight steps. John Kotter published "Leading Change: Why Transformation Efforts Fail" in Harvard Business Review in 1995 and the book Leading Change in 1996, which set out an eight-stage process (Kotter, 1995, 1996). His company now names the steps create a sense of urgency, build a guiding coalition, form a strategic vision, enlist a volunteer army, enable action by removing barriers, generate short-term wins, sustain acceleration and institute change (Kotter International, n.d.). See Kotter's 8-step change model.

Prosci's three phases. Prosci's 3-Phase Process names Prepare Approach, Manage Change and Sustain Outcomes, each with three stages. Prosci describes it as a process for change at the organizational level, while its ADKAR model works at the individual level (Prosci, n.d.).

How it works

A practical five-step version, drawn from the arc the frameworks above share:

1. Prepare. Define the change, who it affects and how much. Name the sponsor and the people who will lead it day to day. Take a baseline of how people are doing now.
2. Plan. Set out how the change will be explained, what training and support people need, what managers need to lead it, and how problems will be raised.
3. Launch and support. Start the change, close to where the work is done where possible, and give managers time and help to carry it with their teams.
4. Review on a rhythm. Meet at a fixed interval to look at what is and is not working, fix what you can, and say what was done about last round's problems.
5. Sustain. Check whether the new way of working is in use, reinforce it, and remove the old ways that compete with it.

The mapping in the table is a reading for comparison, not something the frameworks' authors set out.

Roles. Four roles are commonly described. A sponsor, senior enough to back the change and remove obstacles; Prosci's process includes a sponsor plan (Prosci, n.d.). A change manager or team, who plans the people side. Line managers, who work with their teams day to day; see middle managers. And change agents or champions inside the teams; see change agent and change champions.

Change management and project management. Project management delivers the change itself; change management works on the people who must use it. The two run in parallel, and Prosci says the first phase of its process supports change and project teams (Prosci, n.d.).

What the research says

Frameworks are used loosely. A systematic review of 38 health care studies found 12 change management methodologies in use. Kotter's model (19 studies) and Lewin's (11) were the most common, often used as guiding principles rather than applied step by step (Harrison et al., 2021, rated Medium from its abstract).

How the process is run matters. A systematic review of 10 papers covering 292 health care transformation cases found broad support that dialogic change practice went with a higher likelihood of successful transformation (Hastings, 2025, rated Medium from its abstract). Dialogic practice is the strand of OD described in what is organizational development.

What failed programs lacked. A review of studies from 1980 to 2011 attributed failed change programs to, among other things, insufficient training, inadequate management support, poor communication, inappropriate planning and no monitoring and measurement system (Mosadeghrad & Ansarian, 2014, rated Medium from its abstract).

Step 1 and 2: realistic information. In a two-plant study after a merger announcement, the plant that received a realistic merger preview showed fewer harmful outcomes than the plant that received limited information, and the difference lasted through the study (Schweiger & DeNisi, 1991, rated Medium from its abstract).

Step 3: who starts it. In 1,795 responses from 468 Dutch organizations, employees backed change most when middle managers started it and top managers carried it out, a pattern found in about 3 percent of cases (Heyden et al., 2017, an Atlas study rated medium after a full-text read).

Step 4: debriefs. Debriefs were linked to about 25 percent better performance than comparison conditions across 46 samples (d = .67; .54 with outliers removed), with somewhat smaller effects in real work settings than in simulations (Tannenbaum & Cerasoli, 2013, rated High after a full-text read). After-action reviews worked better when aligned to the individual or team and when they used objective media to review performance (Keiser & Arthur, 2021, rated Medium from its abstract).

Where the evidence is thin

1. No sequence has been tested as a whole. The evidence supports parts of the process, such as realistic communication and structured reviews, not a full step-by-step model against another.
2. Health care. The two reviews of how frameworks are used drew entirely on health care studies.
3. Sustaining. The sources here say little about what keeps a change in place a year or more after launch.

How it maps to the Unfrozen Loop

1. Week 0
   Before you start

   The first step in every framework here is preparation. The Loop's Week 0 adds one thing: measure how people are doing before the change starts, and share the result with them.

2. 1
   Surface both ways

   Step 3 of the process below, supporting people, works better when problems can travel back up. A process with only downward communication leaves out what managers learn on the ground.

3. 2
   Make sense together, top included
4. 3
   Start it close to the work
5. 4
   Review on a rhythm, then hold the pace

   The strongest graded evidence on this page is for debriefs, which were linked to about 25 percent better performance than comparison conditions in a 2013 meta-analysis rated High. Step 4 is where they belong.

6. 5
   The middle’s own weather
7. Ring
   Safe to speak, and speaking changes something

Highlighted parts are the ones this idea speaks to. The mapping is my reading, not a tested finding. See the whole Loop and the evidence for each part.

Common misreadings

"The process is a schedule." Harrison and colleagues (2021) found frameworks often used as guiding principles.

"Change management starts at launch." Preparation and planning come first in every framework above, and the failure review lists inappropriate planning among the reasons programs fail.

"Once it is launched, it is done." Kotter's last steps and Prosci's third phase are about sustaining the change. A missing monitoring and measurement system is on the failure list (Mosadeghrad & Ansarian, 2014).

Using it well

1. Write one page per step. For your next change, write a short page for each of the five steps above. The step you cannot write is the one to work on first.
2. Put the review dates in the calendar before launch. Pick a fixed interval, every two weeks or every month, and hold it even when things seem fine.
3. Give the people side an owner. If no one's job includes the people side of the change, it risks being done by whoever has time.