How to use this plan
This is a template, not a script. Treat each week as a sprint with a defined goal, two parallel workstreams, and a gate check. If you can't pass the gate, do not advance — work the previous week again. Most teams who fail with AI fail because they skip gates and pretend they're further along than they are.
// PHASE 1 · DISCOVERY
Weeks 1–2: See clearly before you move.
No tools, no rollout. Just diagnosis. Most teams skip this and regret it in week 6 when they're scaling something that didn't work.
WEEK 01
Audit & alignment
Goal: one shared picture of where the team is today.
// Working sessions
- Leadership team takes the AI Readiness Audit independently
- 90-min comparison + alignment workshop
- Identify the ONE workflow you'll pilot first
// Async
- Tool inventory: what AI tools is anyone already using?
- Survey team: "what tasks do you do that feel wasteful?"
- Identify the AI champion (the person already excited)
Gate to clear: One named workflow to pilot. One named champion. One agreed-upon "best case after 12 weeks" picture.
WEEK 02
Setup & policy
Goal: tooling sorted, guardrails in place, team briefed.
// Working sessions
- Decide on standard tools (ChatGPT Team or Claude Enterprise default)
- Draft & ratify the AI Use Policy (one page)
- Team all-hands: state the why, the rules, the rollout plan
// Async
- Procurement / IT provisions accounts for the pilot team
- Set up the shared prompt library workspace (Notion / shared doc)
- Block 2 hours of "play time" on every pilot teammate's calendar
Gate: Tools provisioned. Policy in handbook. Pilot team has accounts and protected play time. Team knows what's happening.
// PHASE 2 · PILOT
Weeks 3–4: Prove it with one workflow.
Don't pilot "AI for marketing." Pilot "AI for our weekly content brief." Or "AI for our daily standup notes." Narrow enough that you can measure it. Specific enough that you can iterate.
WEEK 03
Build the first 10 prompts
Goal: ten team-specific prompts that work, saved together.
// Working sessions
- "Prompt sprint": pilot team + champion, 90 min
- For each common task, write a prompt template together
- Test live. Iterate. Save what works.
// Async
- Each pilot member commits to using AI for one daily task
- Daily 60-second async standup: "what AI saved me time today"
- Track hours saved in a shared sheet (just two columns: task + hours)
Gate: 10 saved prompt templates. Hours-saved log started. Daily standups happening.
WEEK 04
Measure & adjust
Goal: real data on whether this is working.
// Working sessions
- End-of-pilot retro (60 min): what's working, what isn't
- Update prompt library with what's been refined
- Decide: expand to next workflow, or fix this one
// Async
- Pilot team writes 3-sentence reflection each
- Calculate hours saved × hourly rate = first ROI number
- Share the ROI number in #ai-wins (or equivalent)
Gate: Documented hours saved per person per week. The pilot delivered ≥ 5 hours saved per person, or you've identified why not and have a fix.
// PHASE 3 · EXPAND
Weeks 5–8: Spread the wins.
Now you have proof from the pilot. Use it. Don't push AI on the next team — pull them with the pilot's story.
WEEK 05
Story-sell the pilot
Goal: every team leader knows what the pilot achieved.
// Working sessions
- 20-min lightning talk at all-hands by pilot champion
- 1:1s with adjacent team leads: "want to be next?"
- Identify next 2 teams to onboard (the eager ones)
// Async
- Write the 1-pager case study (use template in prompt library)
- Share in CEO/founder update or company newsletter
- Pilot team becomes "first responders" for incoming questions
Gate: 2 next teams identified and excited. Case study published internally.
WEEK 06–07
Onboard teams 2 and 3
Goal: two more teams running their own AI workflows.
// Working sessions
- Each new team runs their own 90-min prompt sprint
- Pilot champion attends + coaches
- Each team picks ONE workflow to start with
// Async
- Cross-team prompt library is shared (visible to all)
- Weekly office hours: 30 min, open Q&A on AI
- Hours-saved tracker now multi-team
Gate: Teams 2 and 3 have 5+ saved prompts each. At least one workflow per team showing time saved.
WEEK 08
Surface the resistance
Goal: name what's blocking adoption — before scale amplifies it.
// Working sessions
- 30-min anonymous "what's hard about AI here?" survey
- Leadership reads results, picks top 3 blockers to address
- Public response: acknowledge concerns, name what we're changing
// Async
- Update policy if blockers reveal gaps
- 1:1 with anyone who hasn't logged a single hour saved yet
- Decide: who is the "AI lead" beyond this 90-day plan
Gate: Top 3 blockers named publicly. Plan to address each. AI lead role formalized.
// PHASE 4 · SCALE & MEASURE
Weeks 9–12: Compound the wins.
From "we use AI" to "AI is in our P&L story." The work shifts from rollout to measurement, governance, and habit.
WEEK 09–10
Org-wide adoption push
Goal: every team has at least one active AI workflow.
// Working sessions
- All-team workshop: 90 min, walk through 5 highest-ROI patterns
- Each team commits to one new workflow this month
- AI champion network: one per team, 30-min weekly sync
// Async
- Tracker now company-wide: hours saved per team
- Shared prompt library is a real internal product
- "AI win of the week" goes in CEO newsletter
Gate: Every team has at least 1 active AI workflow. Champion network meeting weekly.
WEEK 11
Measure the P&L impact
Goal: a number leadership can put in the board update.
// Working sessions
- CFO + AI lead: calculate total hours saved × blended rate
- Identify revenue impact: faster sales cycle, better content, etc.
- Build the slide for next board / leadership meeting
// Async
- Audit AI tool spend vs. value delivered
- Decide what to renew, expand, or kill
- Document the framework for Q4 / next year planning
Gate: Quantified ROI number you'd defend in a board meeting. Renew/kill decisions made.
WEEK 12
90-day retro + next 90
Goal: a real retrospective, and a credible plan for the next 90.
// Working sessions
- Full leadership retro (90 min)
- Re-take the readiness audit — measure the delta
- Plan the next 90 days: scale further, deepen current, or pivot
// Async
- Write the public retro: what we tried, what worked, what didn't
- Update the policy with anything we've learned
- Celebrate. Loudly. Tell adoption stories.
Final gate: Audit score increased by at least 5 points. Next 90-day plan ratified by leadership. AI is no longer "a project" — it's "how we work."
// HOW TO CUSTOMIZE
Adapt this to your specific shape.
Small team (5–25 people)
You can compress phases. Pilot in week 2, expand in week 4, scale by week 8. Skip Phase 3's "Onboard teams 2 and 3" — instead, deepen the first workflow.
Mid-size team (25–250 people)
This plan is built for you. Don't compress. Pilot small — one team, one workflow — even if you're tempted to go big.
Large org (250+)
Add a Phase 0 (Weeks -2 to 0): executive sponsorship, IT/security pre-work, legal review of the policy. Don't start Week 1 without these locked.
Regulated industry
Insert governance review at every gate check. Slow Phase 2 down to 4 weeks. Get external counsel review of the AI Use Policy before Week 1.
Want Jay to run this with your team?
Fractional advisory engagements run this exact plan as the spine. 2 slots open.
Book Jay →