Technical how-to

How to do change management communications with AI.

Change management is mostly communication work. New tool rollouts, org changes, policy updates, M&A integration — all require multiple touchpoints across stakeholders. AI compresses the writing without losing the personalization. Here is the framework.

The premise

Why change communications fail

Most change announcements are too sparse (one email, no follow-up) or too generic (same message to every stakeholder group). Both produce confusion, resistance, and rumor cycles.

The version that works: tailored messaging by stakeholder group, multiple touchpoints, clear rationale, anticipated FAQ. AI accelerates the volume; the strategic message stays leadership work.

The 5 communication artifacts

What to create

1. Headline announcement. Same core message across all groups, tailored language.

2. Stakeholder-specific deeper-dives. Why this matters for engineering vs. sales vs. ops.

3. FAQ document. Anticipated questions answered before asked.

4. Manager talking points. What managers should say to their teams in 1:1s.

5. Follow-up communications. Day 7, day 30, day 60 status updates.

The Claude prompt

Use this

I am announcing a change: [WHAT IS CHANGING]

Why we are making this change: [STRATEGIC REASON]
What changes for whom: [STAKEHOLDER GROUP IMPACTS]
Timeline: [SPECIFIC DATES]
Things people will worry about: [LIST]

Generate the 5 communication artifacts:
1. Company-wide headline announcement (200 words)
2. 3-4 stakeholder-specific addendums (per major affected group)
3. FAQ document (10 anticipated questions with honest answers)
4. Manager talking points (1-page)
5. Day 7 follow-up template

Voice rules:
- Direct about the change (do not bury it)
- Honest about uncertainties
- Specific timelines and next steps
- Acknowledge the concerns rather than dismiss them
- No corporate jargon ("synergies", "rightsizing", "rationalization")
Related

Related how-tos

Want change management workflows built?
Implementation includes communications workflow design.
See Implementation → Book the AI Audit