Stage 3: From Insights to Pilots

Part 4 of 7 in the AI Change Framework Series

Assessment clarifies where you stand; Design defines where you're going and how to get there. This stage invites teams to co-create an AI roadmap that connects vision, experimentation, and measurable outcomes.

The Bridge Between Knowing and Doing

Design asks:

• What future do we want AI to enable?
• What must stay the same to protect our identity?
• How can we test bold ideas safely?

Design is not about perfection—it's about clarity of intent and disciplined experimentation that keeps the organization aligned.

Establishing Design Principles

Begin by re-anchoring your AI Purpose Statement from Stage 1. Then ask:

• “What have we learned about our purpose since Assessment?”
• “What principles should guide every AI decision we make?”

Example principles:

• Ethical by design—consider impact before implementation
• Human-in-the-loop—AI augments, not replaces, human judgment
• Measure what matters—define success before starting
• Transparency in automation—people know when AI is involved

Capture 4–6 principles; they'll serve as filters for future pilots.

Opportunity Prioritization (2–3 hours)

1. Start from Stage 2 readiness gaps
2. Build a grid: Impact vs Feasibility
3. Teams brainstorm candidate initiatives and place them on the grid
4. Top-right quadrant = “High impact, high feasibility”—your pilot candidates
5. Select 2–3 to move forward

Pilot Design Sprint (3–4 hours)

For each chosen pilot, define:

• Problem Statement—What specific pain are we solving?
• Goal—What does success look like?
• Metrics—How will we measure progress?
• Stakeholders—Who needs to be involved?
• Risks—What could go wrong?
• Timeline & Resources—What do we need?

“The facilitator ensures every pilot ties back to the AI Purpose Statement. Avoid random acts of automation.”

Governance & Learning Loops

Define how pilots will be tracked, reviewed, and communicated:

• Meeting cadence—e.g., bi-weekly stand-up + monthly steering review
• Reporting method—shared dashboard or brief narrative
• Decision checkpoints—continue, pivot, or stop

Learning Loop Structure

Decide how learning will be captured and shared:

Hypothesis → Action → Result → Learning → Next step

Assign a “learning owner” for each pilot.

Mastery Checklist

The Design stage is complete when:

1. A clear AI Vision Narrative and Design Principles are approved
2. A portfolio of prioritized, resourced pilots exists
3. Governance, communication, and feedback mechanisms are operational
4. Teams understand how success will be measured
5. Leadership publicly commits to experimentation as a core practice

Artifacts to Produce

• ☐ AI Vision Narrative completed
• ☐ Design Principles agreed upon
• ☐ Pilot Charters drafted and approved
• ☐ Governance and Comms Plan in place
• ☐ Learning loop structure defined

Previous: Stage 2: Diagnosing AI Readiness
Next: Stage 4: Launching AI with Learning Loops