Change Management
Framework for rolling out organizational changes without chaos.
$skillfish add ~skills/change-management Real Conversations
See how the Change Management skill responds to actual requests. These are representative examples of what your AI agent will produce.
We've never had formal performance reviews in 4 years. Now at 38 people we're implementing them. Engineering team is resistant — they see it as bureaucracy. HR says we need it for promotions and comp equity. Rollout is in 6 weeks. How do I do this without a culture revolt?
Engineering resistance to performance reviews is almost always about fear, not logistics — fear of arbitrary evaluation, reduced autonomy, or "being managed." Address the fear directly.
ADKAR Framework Applied:
A — Awareness (Week 1) Don't announce "we're doing performance reviews." Announce the problem you're solving: "We have 8 people who deserve promotions and no consistent process to evaluate them fairly. We have comp decisions being made inconsistently. We need a system."
D — Desire (Week 1-2) The engineering team will engage if they see benefit to themselves:
- Engineers who are performing well WANT a process that recognizes them objectively
- Identify your 3-4 top engineering performers and brief them privately first — their buy-in is your social proof
K — Knowledge (Week 2-3) Show them the process before it happens to them:
- Share the review template publicly 4 weeks before first cycle
- Hold a Q&A session: "What would make this useful vs. useless?"
- Incorporate 2-3 engineering-specific pieces of feedback
A — Ability (Week 3-5) Train managers first. The #1 failure mode is managers who don't know how to give feedback — engineers will blame the process when the real problem is manager skill.
R — Reinforcement (Week 6+) First promotion or comp adjustment that comes from the new process = proof of concept. Make that story public.
What to avoid: Launching with a heavy form. Start with a 3-question self-assessment + 30-minute 1:1. Complexity can come in cycle 2.
What's driving the 6-week deadline — is there a comp cycle or fundraise attached to this?
Get Started
How to install and use this skill in your preferred environment.
Skills are designed for AI coding agents (Claude Code, Cursor, Windsurf) and IDE-based workflows where the agent can read files, run scripts, and act on your codebase. Web-based AI can use the knowledge and frameworks, but won't have tool access.
Models & Context
Which AI models and context windows work best with this skill.
Recommended Models
Larger models produce more detailed, production-ready outputs.
Context Window
This skill's SKILL.md is typically 3–10 KB — fits in any modern context window.
All current frontier models (Claude, GPT, Gemini) support 100K+ context. Use the full window for complex multi-service work.
Pro tips for best results
Be specific
Include numbers — users, budget, RPS — so the skill can size the architecture.
Share constraints
Compliance needs, team size, and existing stack all improve the output.
Iterate
Start with a high-level design, then ask follow-ups for IaC, cost analysis, or security review.
Combine skills
Pair with companion skills below for end-to-end coverage.
Good to Know
Advanced guide and reference material for Change Management. Background, edge cases, and patterns worth understanding.
Contents
ADKAR Model Deep-Dive
ADKAR describes the five building blocks an individual must have for change to stick. A person can fail at any element, and the failure mode is different for each.
| Element | What It Means | Symptom When Missing | Intervention |
|---|---|---|---|
| Awareness | Understanding why the change is happening | People ask "why are we doing this?" or assume the worst | Leadership communication explaining the business reason — not the feature list |
| Desire | Personal motivation to support and participate | Compliance without engagement; passive resistance | Address WIIFM ("what's in it for me"); involve influencers; surface and resolve fears |
| Knowledge | Knowing how to change (skills, process, tools) | People want to change but don't know how; mistakes increase | Training, job aids, peer coaching, sandbox environments |
| Ability | Actually being able to perform the new behaviors | Knowledge exists but performance doesn't follow | Supervised practice, coaching, feedback loops, removing environmental barriers |
| Reinforcement | Mechanisms that sustain the change | Reversion to old behavior 3–6 months post-launch | Recognition, accountability structures, removing the old way as an option |
Critical insight: ADKAR is a diagnostic model, not a checklist. When a change is failing, identify which element is the weakest link — don't restart from Awareness if the problem is Ability.
Resistance Taxonomy
| Archetype | How to Identify | Engagement Approach |
|---|---|---|
| Skeptic | Asks hard questions publicly; intellectually engaged but unconvinced | Provide data and evidence; involve them in design — skeptics become champions when they feel heard |
| Saboteur | Appears compliant but undermines the change in private | 1:1 direct conversation; name the behavior, not the person; escalate if behavior continues |
| Silent Non-adopter | Attends training, says nothing, reverts to old behavior quietly | Proactive check-ins; peer buddy pairing; make non-adoption visible through metrics |
| Vocal Objector | Raises objections loudly in group settings | Acknowledge their concern publicly, then resolve privately; give them a formal feedback channel |
| Overloaded Adopter | Wants to change but has too many concurrent demands | Audit their change load; defer lower-priority changes; provide dedicated time/resources |
| Late Follower | Waits to see if the change sticks before committing | Share early adopter success stories; reduce perceived risk of adoption |
Kotter's 8-Step vs ADKAR
| Dimension | Kotter's 8 Steps | ADKAR |
|---|---|---|
| Unit of analysis | The organization | The individual |
| Primary use case | Large-scale org transformation | Adoption measurement and coaching |
| Output | Transformed organization culture | Individuals who have successfully changed |
| Where it breaks down | Doesn't tell you why specific individuals aren't adopting | Doesn't provide an org-level change architecture |
When to use Kotter: You are redesigning an organizational structure, merging two cultures, or executing a transformation that requires building a new coalition of leadership. Kotter's strength is in creating and sustaining organizational momentum.
When to use ADKAR: You need to know why adoption is low after a rollout, or you're coaching a manager on how to bring their team through a change. ADKAR is operational; Kotter is strategic.
In practice, use both: Kotter to design the program, ADKAR to diagnose and course-correct at the individual level.
Communication Timing
The "7 touches" rule: Research consistently shows people need to encounter a change message 5–7 times before it registers. This is not redundancy — it is how human cognition works. Plan your communication calendar with this as a constraint, not an inconvenience.
Framing by audience:
- Executives: Business rationale, risk of inaction, timeline, resource requirements. Two paragraphs maximum.
- Managers: What they need to tell their teams, what questions to expect, what resources are available, what success looks like for their function.
- Individual contributors: What changes for them specifically, what stays the same, when it happens, where to go with questions.
Change communication calendar structure: Map communications across three phases — Before (building awareness and desire), During (enabling knowledge and ability), and After (reinforcing and sustaining). Each phase needs different channels: email for formal announcements, team meetings for dialogue, manager conversations for personal concerns.
Change Saturation
Organizations have a finite capacity to absorb change. When too many initiatives run simultaneously, adoption rates fall across all of them — not just the newest ones.
Detection signals: Employees cite "another change" with fatigue rather than curiosity; training completion rates drop; adoption metrics plateau at 40–60% and don't improve; managers start selectively enforcing which changes they champion.
Change portfolio audit: Inventory all active change initiatives with their status, affected populations, and estimated adoption effort. Map them against the affected teams to identify where single groups are being hit by 3 or more concurrent changes.
Sequencing approach: Prioritize changes by strategic value and adoption complexity. Defer or pause lower-priority changes until bandwidth exists. High-complexity changes (new systems, structural reorganizations) should not run in parallel with high-frequency changes (process updates, policy rollouts) for the same population.
Ready to try Change Management?
Install the skill and start getting expert-level guidance in your workflow — any agent, any IDE.
$skillfish add ~skills/change-management