What Are the 7 Key Steps of Change Management?
The 7 key steps of change management are: create urgency, build a guiding coalition, form a strategic vision, enlist a volunteer army, enable action by removing barriers, generate short-term wins, and sustain acceleration to anchor change in culture. This framework, developed by John Kotter, remains the operational reference for any leader driving deep organizational transformation.
Why These 7 Steps?
There are dozens of change management models. Lewin, ADKAR, Prosci. Each has its advocates.
But in the projects I work on, particularly AI deployments in organizations, Kotter’s 7-step model remains the most operational. Not because it’s perfect. Because it follows the human logic of change: you don’t convince people with a PowerPoint. You convince them by creating a context where not changing becomes riskier than changing.
Here’s how to apply each step concretely.
Step 1: Create a Sense of Urgency
Without perceived urgency, nothing moves. Teams wait. Managers stall.
Urgency isn’t decreed. It’s built. Show what’s happening at your competitors. Show what you’re losing every quarter by not acting. In Morocco, the launch of Gemini Enterprise by Maroc Cloud puts concrete pressure on executive teams: enterprise-grade AI tools like Gemini Enterprise are now available locally. The argument “we’ll see in two years” no longer holds.
Step 2: Build a Guiding Coalition
A project carried by one person, even the CEO, fails. You need a coalition: visible sponsors at different levels of the organization.
In an AI deployment project, this coalition typically includes the CHRO, CIO, a business unit director, and ideally a board member. Not for organizational chart aesthetics. So that every resistance meets a legitimate interlocutor.
Step 3: Form a Strategic Vision
The vision must fit in three sentences. If you can’t explain it to a frontline manager in under two minutes, it’s too complex.
Concrete example: “Within 18 months, our recruiters spend less time screening CVs and more time on human candidate assessment. AI handles volume. Our teams handle decisions.”
That’s precise. Human. Sellable internally.
If you’re looking to structure this vision in an AI integration context, my practical guide for leaders details the questions to ask before launching anything.
Step 4: Communicate the Vision at Scale
Most leaders communicate once, at launch. Then they’re surprised that teams haven’t internalized the message.
Kotter recommends multiplying channels and repetitions. Team meetings, internal memos, one-on-ones with key managers, concrete examples at executive committees. The vision must be present everywhere, not just in the launch email.
I’ve built a 6-dimension diagnostic framework to assess an organization’s change readiness, including internal communication quality. Download the AI Board Pack 2026 to use it directly with your leadership team.
Step 5: Remove Obstacles
This is the step leaders underestimate most.
You’ve communicated. Teams understand the direction. But nothing moves. Why? Because there are concrete obstacles: an incompatible IT system, a middle manager who’s blocking, an HR policy that penalizes risk-taking, nonexistent training.
Identify these obstacles one by one. Address them. Don’t work around them.
In AI adoption projects, the most frequent obstacle I observe is the absence of AI literacy among frontline managers. They don’t know what to tell their teams. They’re afraid of appearing out of touch. That’s not ideological resistance. It’s discomfort with the unknown.
Step 6: Generate Short-Term Wins
An 18-month project with no visible result at 3 months is a dead project.
Plan quick wins from the start. Not cosmetic wins. Measurable results, even modest ones, that prove the direction is right.
In an AI-assisted recruitment project, a quick win might be: reduced pre-screening time on a pilot role, positive manager feedback on candidate quality. As I explained in my analysis on integrating AI into recruitment, starting small and measuring fast is the only way to build internal trust.
Step 7: Anchor Change in Culture
This is the final step, and the longest.
Change isn’t anchored until new practices are embedded in evaluation, hiring, and promotion processes. If you deploy AI in your HR processes but your manager evaluation criteria haven’t changed, the change won’t hold.
Anchoring means connecting new practices to achieved results. “Here’s what we accomplished. Here’s why we’re not going back.”
What These 7 Steps Change in an AI Project
The Moroccan context is particular. Companies are receiving strong signals: Al Akhawayn is preparing its graduates for new AI applications, with the university itself noting that AI transforms their missions rather than eliminating their jobs. Morocco has partnered with Vertiv to build AI-dedicated infrastructure. Higher education now counts 80 engineering and master’s programs in AI. The ecosystem is structuring quickly.
But enterprise adoption doesn’t automatically follow technological supply. That’s where change management makes the difference. The tools exist. The question is whether your teams are ready to use them, and whether your organization is structured to generate measurable value from them.
For a deeper look at what AI actually changes in organizations, read my analysis on AI’s real role in business.
If you’re a CHRO or CEO looking to structure your change management approach around an AI project, request a free diagnostic. We’ll look together at where you stand on each of the 7 steps.
FAQ
What’s the difference between Kotter’s model and ADKAR?
Kotter is sequential and organizational: it describes what the organization must do, step by step. ADKAR is individual: it describes what each person must go through (Awareness, Desire, Knowledge, Ability, Reinforcement). Both are complementary. Kotter drives the project. ADKAR drives the person.
How long does successful change management take?
There’s no standard duration. A limited process change can stabilize in a few months. A deep operational model transformation often takes 18 to 36 months before it’s truly anchored. What matters is not confusing the launch with the anchoring.
What are the concrete tools of change management?
Among the most used tools: stakeholder mapping, resistance analysis, communication plans, adoption tracking dashboards, and lessons-learned sessions. The tool isn’t the method. It serves the method.
How do you manage resistance to change in an AI project?
First, understand where it comes from. Resistance is rarely ideological. It’s often linked to fear of losing competence, status, or relevance. The answer isn’t to force. It’s to involve early, train, and show that AI redefines missions without eliminating them, as Al Akhawayn’s own observations about its graduates demonstrate.