12-Week AI Adoption Guidebook

Guidebook Use

• This workbook is a 12-week adoption guide: one 45-minute session per tab.
• It is recommended that you download the XLSX version of this guidebook here
• Leaders: skim each weekly tab before your session. The ‘Instructor Directions’ column is your playbook to shine.
• Capture live notes in the Feedback tab. Paste selected rows into the Prompts to generate summaries and emails.
• Keep scope small. Celebrate learning, not perfection. Adoption is a rhythm, not a rollout.

Quick Tips

• Model curiosity: share one imperfect experiment first.
• Protect time: if playtime isn’t scheduled, it won’t happen.
• Name risks early; define controls together (reviewers, audit, rollback).
• Credit skeptics who contribute — it builds durable trust.

Week 1

Session Theme

Kickoff: Setting the Stage for AI Adoption

Objectives

• Align on why we're exploring AI now (change management over tech rollout)

• Normalize curiosity and questions; set tone of ‘learn together’

• Surface one real work area per person that could benefit from AI

Agenda (45 min)

5 min — Welcome & context: Why adoption is a people project

10 min — Live demo: Simple, relatable AI use (e.g., summarizing a policy)

15 min — Round-robin: One task you'd make easier or faster

10 min — Agree on ground rules (psych safety, data do/don’t, sandbox)

5 min — Assign light experiment: Try one AI interaction before Week 2

Materials

• Slides (optional) with ground rules and safe-use policy highlights

• Live access to an approved AI tool

• Note-capture doc or shared spreadsheet (use Feedback tab)

Instructor Directions

• Model humility: Share one thing AI confuses you about

• Co-create the ground rules; don't read them at people

• Invite one-word reactions to the demo to take pulse (e.g., 'curious', 'skeptical')

Discussion Prompts

• What part of today’s work would you happily outsource to a ‘smart intern’?

• What’s one outcome (not output) that AI might help us achieve?

• What boundary do we need so people feel safe experimenting?

Risk to Watch

• Over-promising on efficiency before trust is built

• Data copy/paste into tools without clarity on sensitivity

• Equating curiosity with commitment — keep pressure low

Success Criteria

• Ground rules agreed and posted

• Each participant named a candidate task to test

• Sentiment check recorded (e.g., word cloud from reactions)

Follow-ups

• Share the ground rules and resources link

• Confirm one micro-experiment per person before Week 2

Week 2

Session Theme

Exploring Everyday AI Use Cases

Objectives

• Showcase simple tools and low-risk workflows

• Capture first impressions (wins, frictions, surprises)

• Identify 1-2 department-specific ‘use case clusters’

Agenda (45 min)

5 min — Check-in: What did you try? One-sentence share

10 min — Demo: Drafting email, summarizing a page, synthesizing notes

15 min — Breakouts by function: Map 3 candidate tasks per group

10 min — Report-out and cluster themes

5 min — Decide which tasks to test in sandbox (Weeks 5–8)

Materials

• Short list of approved tools (with links)

• Use-case cards or examples from your domain

• Feedback tab ready for notes

Instructor Directions

• Capture ‘frictions’ explicitly — they inform guardrails

• Praise skepticism that is specific (turns into design input)

• Keep demos short, interactive, and relevant

Discussion Prompts

• Where did AI feel helpful vs. distracting?

• What would make your next experiment safer or clearer?

• Which use cases seem shareable across teams?

Risk to Watch

• Performative buy-in (nods without engagement)

• Tool sprawl without clarity on data boundaries

Success Criteria

• 2–3 use case clusters identified per department

• Specific frictions documented to inform policy

Follow-ups

• Publish the ‘shortlist’ of sandbox tasks for Month 2

• Assign owners for each task

Week 3

Session Theme

Understanding Stakeholder Perspectives

Objectives

• Map roles most affected by AI changes

• Elicit concerns and incentives per role

• Draft first-pass personas (with ChatGPT) for adoption design

Agenda (45 min)

5 min — Recap: Where AI might touch identity, status, or risk

15 min — Live persona creation in ChatGPT (role-by-role)

15 min — Group refine: What would build trust for each persona?

10 min — Decide on outreach/enablement tactics per role

Materials

• Role list by department

• ChatGPT prompt template (see Prompts tab)

Instructor Directions

• Name the elephant: Some roles feel automated away — center trust

• Translate ‘fear’ into design requirements (e.g., manual override, auditability)

Discussion Prompts

• What would make this feel like a tool rather than a threat?

• What decision rights must remain human?

• What ‘proof of safety’ would each role need?

Risk to Watch

• Stereotyping personas — keep them grounded in your org’s reality

• Skipping enablement for skeptics (they need *more*, not less)

Success Criteria

• Personas drafted with clear trust-builders

• Enablement tactics aligned to roles

Follow-ups

• Share personas and invite async comments

• Fold insights into sandbox design

Week 4

Session Theme

Building a Culture of Curiosity

Objectives

• Codify shared principles for experimentation

• Make leader behaviors explicit (model curiosity, protect time)

• Prepare for Month 2 sandbox launch

Agenda (45 min)

10 min — Principles workshop: Draft 5–7 team norms

15 min — Leader modeling: Show a messy experiment and learnings

10 min — Define the sandbox rituals (cadence, channels, learning log)

10 min — Confirm Month 2 plan and owners

Materials

• Principles examples

• Learning log template (use Feedback tab or your wiki)

Instructor Directions

• Choose *one* visible leader habit to start this week (e.g., ‘Friday Fumbles’)

• Timebox decisions; perfect is the enemy of momentum

Discussion Prompts

• What ritual will make curiosity durable here?

• Where do we need a ‘do not use AI’ sign for now (compliance/data)?

Risk to Watch

• Principles that are posters, not practices — assign owners

• No time protected for play — sandbox dies on impact

Success Criteria

• Principles finalized and published

• Sandbox cadence and channels created

Follow-ups

• Announce Month 2 sandbox and expectations

• Remind leaders of their modeling commitment

Week 5

Session Theme

Launching the AI Playground (Sandbox)

Objectives

• Clarify boundaries, approved tools, and success measures

• Kick off low-stakes experiments tied to real tasks

• Start the learning log habit

Agenda (45 min)

10 min — Policy recap: Data rules, do/don’t examples

15 min — Team selects 1–2 experiments from shortlist

15 min — Plan experiments: who, what, how, evidence of learning

5 min — Commit in writing to Feedback tab

Materials

• Approved tool list with links

• Experiment planning template (use Feedback tab columns)

Instructor Directions

• Ask for ‘evidence of learning’ not ‘proof of ROI’

• Keep scope tiny; aim for momentum over magnitude

Discussion Prompts

• What’s the smallest step that would teach us the most?

• How will we know we learned something useful?

Risk to Watch

• Shadow data use; re-iterate safe examples

• Overengineering experiments

Success Criteria

• At least one experiment per team committed

• Learning log entries started

Follow-ups

• DM owners mid-week to unblock

• Capture quick video wins to share

Week 6

Session Theme

Experimentation Round 1

Objectives

• Share early results (warts and all)

• Distill surprises and blockers

• Refine experiments for next week

Agenda (45 min)

5 min — Lightning updates (60s each)

15 min — Show & tell: before/after artifacts

15 min — Group debrief: patterns, blockers, workarounds

10 min — Adjust next tests; log changes

Materials

• Screenshots or snippets of outputs

• Feedback tab ready to capture patterns

Instructor Directions

• Normalize ‘failed’ experiments as progress

• Extract one reusable prompt per team

Discussion Prompts

• What would you try differently with the same task?

• Where did human judgment add essential value?

Risk to Watch

• Tool-of-the-week distractions

• Judgmental language that chills participation

Success Criteria

• Blockers identified with owners

• At least one re-usable prompt captured

Follow-ups

• Share a mini ‘pattern library’ of prompts

• Check compliance on any gray-area data

Week 7

Session Theme

Experimentation Round 2 + Peer Sharing

Objectives

• Deepen experiments and cross-pollinate ideas

• Pair teams to give feedback

• Capture candidate pilots

Agenda (45 min)

10 min — Pairs exchange: demo & critique

15 min — Refactor prompts or workflow live

10 min — Identify what’s pilot-worthy vs. playground-only

10 min — Update Feedback tab with candidate pilots

Materials

• Role list by department

• ChatGPT prompt template (see Prompts tab)

Instructor Directions

• Keep critique kind, specific, and actionable

• Ask ‘what problem did this actually solve?’

Discussion Prompts

• What made this transfer well to a different team?

• What trust checks would a pilot need (QA, audit, reviewers)?

Risk to Watch

• Premature scaling without trust gates

• Overfitting prompts to one person’s style

Success Criteria

• 2–3 pilot candidates listed with rationale

• Revised prompts documented

Follow-ups

• Draft pilot charters for top candidates

• Confirm reviewers and success gates

Week 8

Session Theme

Identifying Early Wins

Objectives

• Surface credible small wins (confidence builders)

• Decide which wins to communicate org-wide

• Prepare transition from sandbox to practice

Agenda (45 min)

10 min — Win stories: what changed for real people

10 min — Evidence review: before/after artifacts

15 min — Comms plan: who needs to hear what

10 min — Shortlist 1–2 pilots to ready for Month 3

Materials

• Artifacts, quotes, screenshots

• Comms template (Prompts tab)

Instructor Directions

• Spotlight contributions by skeptics who engaged

• Translate wins into human outcomes (clarity, speed, confidence)

Discussion Prompts

• Which story would persuade a thoughtful skeptic?

• What caveats must be shared to stay credible?

Risk to Watch

• Over-claiming impact from thin evidence

• Skipping attribution; give credit generously

Success Criteria

• Win stories drafted and approved

• Pilot shortlist confirmed

Follow-ups

• Publish internal post with wins

• Set up pilot kickoffs for Month 3

Week 9

Session Theme

From Play to Practice

Objectives

• Select experiments ready for real workflows

• Define decision rights, QA, and oversight

• Plan lightweight adoption in one area

Agenda (45 min)

10 min — Pilot candidate deep-dive

15 min — Define controls: reviewers, audit trails, rollbacks

10 min — Decide where to embed first

10 min — Document practice change in Feedback tab

Materials

• Pilot charter template

• Risk/controls checklist

Instructor Directions

• Insist on explicit human-in-the-loop steps

• Name exit criteria if the pilot underperforms

Discussion Prompts

• What could go wrong and how will we catch it?

• What training is required for the target workflow?

Risk to Watch

• Ambiguous ownership for QA

• No rollback plan

Success Criteria

• Pilot plan with owners and controls

• Training needs logged

Follow-ups

• Book training sessions

• Prepare success/stop criteria doc

Week 10

Session Theme

Cross-Team Learning

Objectives

• Share pilots-in-progress to harvest patterns

• Reduce duplication by aligning methods

• Spot opportunities for shared components

Agenda (45 min)

10 min — Pilot posters: 3-minute per team

15 min — Pattern harvest: prompts, workflows, controls

10 min — Consolidate into shared ‘starter kits’

10 min — Assign maintainers

Materials

• Poster template

• Starter kit repo/folder

Instructor Directions

• Ask ‘What can we standardize without stifling learning?’

• Capture 3–5 reusable assets

Discussion Prompts

• Which patterns are generalizable vs. context-specific?

• What enablement would let another team adopt this next week?

Risk to Watch

• Inconsistent controls across teams

• Starter kits that no one maintains

Success Criteria

• Shared assets published

• Maintainers named

Follow-ups

• Announce starter kits and where to find them

• Schedule office hours

Week 11

Session Theme

Measuring Progress (Beyond Efficiency)

Objectives

• Define qualitative & lightweight quantitative signals

• Assess confidence and collaboration shifts

• Decide on ongoing measurement cadence

Agenda (45 min)

10 min — Review learning log & pilot outcomes

15 min — Define 4–6 signals (e.g., time-to-draft, confidence rating)

10 min — Agree how to capture signals monthly

10 min — Draft scorecard outline

Materials

• Signal/metric menu

• Scorecard template

Instructor Directions

• Keep measures human-centered and honest

• Avoid vanity metrics that invite gaming

Discussion Prompts

• What feels different about your work now?

• Where is oversight still needed?

Risk to Watch

• Chasing ROI prematurely

• Survey fatigue

Success Criteria

• Signals selected and documented

• Scorecard skeleton created

Follow-ups

• Publish scorecard and owners

• Schedule first readout

Week 12

Session Theme

Reflect, Celebrate, and Plan Next Steps

Objectives

• Consolidate learnings into a narrative

• Decide what becomes ‘business as usual’

• Set priorities for the next 90 days

Agenda (45 min)

10 min — Celebrations and shout-outs

15 min — Story of change: Before → After → Next

10 min — BAU candidates vs. retire/reshape

10 min — Top 3 priorities for next quarter

Materials

• Learning log highlights

• Comms templates (Prompts tab)

Instructor Directions

• Make heroes of contributors (especially early skeptics)

• Keep ‘next’ small, specific, and scheduled

Discussion Prompts

• What stays, what evolves, what stops?

• What support do we need to keep momentum?

Risk to Watch

• Letting the story die in files — tell it widely

• No owner for BAU changes

Success Criteria

• Change story drafted

• Next 90-day priorities assigned

Follow-ups

• Send summary email & publish artifacts

• Book next quarter’s cadence

Feedback Template

How to use

• Log notes live during or after each session. Keep items concrete and attributable.
• Paste links to artifacts (screens, prompts, docs). Include brief context.
• Capture at least one quote per week in participants’ own words.
• Use ‘Confidence’ as a quick pulse. Low scores are valuable signals.
• ‘Next Step’ should be specific (owner + date).

Use these columns (or download the .XLSX file above)

• Week
• Team/Function
• Owner
• Experiment/Topic
• Wins (What helped)
• Challenges (What hindered)
• Surprises/Insights
• Quality & Trust Notes
• Risks/Controls Needed
• Artifacts/Links
• Quotes (Participant Voice)
• Metrics/Signals
• Confidence (1-5)
• Support Needed
• Next Step & By When

Prompts

Weekly Synthesis (Paste Feedback Rows)

You are an AI adoption coach. Synthesize this week’s entries from our Feedback table (pasted below). Create: (1) 5–7 bullet insights (wins, frictions, surprises), (2) a short risk/controls note, (3) 3 recommended experiments for next week with smallest viable scope, (4) one leader spotlight recognizing a specific contribution.

Feedback rows:

[[PASTE ROWS HERE]]

Executive One-Pager

Draft a crisp executive one-pager summarizing AI adoption progress for Week {{WEEK}}: Context, Highlights, Risks & Mitigations, Next Week Focus. Use plain language, 150–200 words, bullet-friendly.

Follow-up Email to Participants

Write a concise follow-up email to participants recapping Week {{WEEK}}. Sections: Thank you, What we learned (3 bullets), Decisions made, What to try before next session, How to ask for help. Friendly, pragmatic tone. 150–220 words. Subject line: ‘AI Sandbox Week {{WEEK}} — What’s Next’.

Pilot Candidate Selection

From the current Feedback data, propose 2–3 pilot candidates. For each: problem statement, expected benefit, minimal viable guardrails (reviewers, audit steps), and exit criteria. Keep each pilot description under 120 words.

Stakeholder Personas (Trust Builders)

Generate stakeholder personas for roles listed: {{ROLES}}. For each persona: motivation, specific concerns, trust builders (what they need to see/experience), and a 2-sentence engagement tactic.

Change Story (Before → After → Next)

Write a narrative of change: Before → After → Next, using highlights from the Feedback. Make it people-centered, credit contributors, and end with 3 specific next actions.