What Are the 3 Phases of Change Management?
The 3 phases of change management are preparation, deployment, and anchoring. First, you explain why the change is needed and what will change in practice. Then, you support teams through adoption. Finally, you stabilize the new habits so the organization does not slide back. Without these three steps, change management in business stays superficial.
Why this question matters
In a company, the problem is rarely the idea. The problem is execution. A new system, a new organization, a new tool, all of it often fails in the same place. Teams do not understand, managers do not relay the message, and leadership assumes the launch was enough.
That is why the phases of change management must be treated as a management sequence, not as a communication campaign.
When I see companies talking about AI, process redesign, or a new operating model, I always ask the same question. Who carries the change day to day? If the answer is vague, the project is vague too. I also cover this in my analysis on how to integrate AI in a company.
Phase 1. Prepare the ground
The first phase is about clarity. Why are we changing? What is no longer working? What will change for teams, managers, and clients?
This is where many leaders lose time. They talk about vision, but not about operational impact. Yet teams do not buy into an abstract vision. They buy into a clear explanation of what changes in their daily work.
In this phase, you also need to identify key stakeholders. Not only sponsors. Frontline managers. Support functions. The people who will absorb questions, resistance, and doubt.
Lewin’s model is still useful here. You “unfreeze” existing habits before asking for a new effort. Without that work, you ask people to move without explaining why they should leave their comfort zone.
What a manager should do
- State the problem in simple terms.
- Explain the expected gain for the business and for teams.
- Map the impact by role.
- Prepare the core messages, not just the slides.
If you want to structure this properly, I often include it in a methodological framework for AI governance, because the same discipline applies to any major change.
Phase 2. Deploy with discipline
The second phase is where action starts. This is when change management becomes visible. You train, support, correct, listen. This is the moment when promises meet the field.
Kotter’s model is useful to understand this stage. It emphasizes urgency, the guiding coalition, repeated communication, and quick wins. On the ground, that means one thing. You need traction.
A launch without support creates confusion. A launch with trained managers, consistent messages, and regular feedback creates buy-in.
Take a concrete example. A services company rolling out a new sales tracking tool should not only train teams on where to click. It must also redesign routines, indicators, responsibilities, and decisions. Otherwise, the old system comes back through the back door.
This is also where companies underestimate the mental load of change. Teams must learn a new tool, but they also have to drop old habits. That is costly. It should be said plainly.
What a manager should do
- Give short, repeated instructions.
- Organize frequent follow-up meetings.
- Surface friction early.
- Highlight visible early wins.
Phase 3. Anchor it for the long term
The third phase is the most neglected. Yet it is the one that decides the real outcome. A change is not successful when it is launched. It is successful when it becomes normal.
Anchoring happens through routines, indicators, evaluations, and managerial behavior. If old habits are still rewarded, the change will retreat. If new behaviors are built into goals and rituals, they will stick.
In a company, that may mean revising job descriptions, evaluation criteria, management reports, and steering meetings. Not for show. To make the change irreversible.
This is also the moment when leadership must look at the gap between words and practice. If a manager keeps rewarding the old way of working, the project is already weakened.
The mistakes I see most often
The first mistake is believing communication is enough. It is not. Communication opens the door, it does not get people across the room.
The second mistake is forgetting frontline managers. They translate strategy into daily actions. If they are not involved, the change stays theoretical.
The third mistake is not measuring adoption. Companies measure the launch, but rarely the real usage. That is a management failure.
In AI-related projects, this mistake is even more visible. Companies talk about tools, but not about responsibility and accountability, or guardrails. That is why I detailed the main risks in my article on the 3 drawbacks of AI.
How to apply these 3 phases without overcomplicating things
Do not turn change management into a bureaucracy machine. Three questions are often enough to stay on track.
- What exactly is changing?
- Who needs to do things differently, and when?
- How will we know the change is truly sticking?
If you can answer those three questions clearly, you already have a solid structure. If you cannot, you mostly have a communication plan.
For leaders who want to go further, I have gathered the key watchpoints in my services and on my about page, because change management is never just theory. It is a matter of execution.
Conclusion
The 3 phases of change management are simple to name, but demanding to execute. Prepare, deploy, anchor. That is the minimum sequence to avoid a project fading after launch.