Format: A 30-minute client briefing. 50-minute session, shared time slot. Brandon has 30 minutes: 25 minutes of content, 5 minutes of Q&A. Every minute must earn its place.
Pre-Workshop (72 Hours Before)
Send attendees the Executive One-Pager (findings/executive-one-pager.md) as a single-page PDF. No slide deck. Frame it: “A 3-minute read to get the most from Thursday’s session.”
Optionally send the Assessment Tool link: “Takes 10 minutes. Bring your score if you’d like to benchmark against the room.”
[INTERNAL] Arrive early. Shake hands. Learn names. Ask what they’re hoping to get out of the session. The real consulting starts before you present.
Materials
| Item | Purpose | Format |
|---|---|---|
| Executive One-Pager | Pre-read + leave-behind | Printed, 1 page |
| AI Native Adoption Cycle diagram | Visual anchor, handout | Printed, 1 page |
| 90-Day Quick Start | Post-session follow-up | Digital, emailed after |
No slide deck over 12 slides. Heavy on data visuals, zero slides about Brandon.
The 30 Minutes
0:00–0:02 — Open Cold
“Five percent of companies capture real value from AI. Ninety-five percent don’t. The difference is not the technology — everyone has access to the same tools. In the next 25 minutes, I’ll show you what separates them and give you a framework to act on Monday morning.”
No bio. No “a little about me.” The content is the credibility.
[INTERNAL] Data: BCG (n=1,250+) and McKinsey (n=1,993) independently converge on 5-6%. Two surveys, two methodologies, same answer.
0:02–0:08 — What Works (The Proof)
Lead with named wins — 2 minutes:
| Company | What They Did | Result | Duration |
|---|---|---|---|
| UPS | ML route optimization | $400M/yr savings | 10+ years |
| JPMorgan | Fraud detection (400+ AI use cases) | $1.5B prevented annually | Sustained |
| Citi | 4,000 peer champions, no mandates | 70% adoption, 182K employees | 2 years |
| IKEA | Reskilled 8,500 workers (not layoffs) | $1.4B new revenue | Ongoing |
“These are not pilot results. UPS has been running this for a decade. The question is not whether AI works. It’s whether your organization is structured to capture it.”
Where it works task by task — 2 minutes:
Tier 1 (deploy now, proven): Code autocomplete, test generation, documentation, meeting summaries, customer service routing — 25-35% gains, all configuration, no custom development.
Tier 2 (deploy in 3-9 months): AI code review, contract analysis, financial forecasting, knowledge base Q&A — requires workflow changes.
“The boring AI is the profitable AI. Invoice processing, code autocomplete, and customer service routing deliver measurable gains within 90 days (Atlan, 200 deployments, 2025). The glamorous stuff — autonomous agents, AI architects — is where the hype lives, not where the ROI lives.”
What the 5% do differently — 2 minutes:
Three things:
- Deploy on the right tasks first. Not everything at once. Tier 1 tasks, then build.
- Budget for the real cost. License fees are 4-9% of actual cost. Training, governance, workflow redesign are the other 91-96%. BCG’s rule: 70% of investment on people and process, 20% on technology, 10% on algorithms.
- Redesign around the new bottleneck. Faros AI telemetry (n=10,000+ developers, 1,255 teams, 2025) shows 98% more PRs merged but 91% longer review times and zero improvement in DORA delivery metrics. The speed evaporates unless you address where the constraint moved.
[INTERNAL] Supporting: findings/what-actually-works.md, findings/proven-ai-case-studies.md, findings/ai-change-management-playbook.md. Do NOT name-drop consulting firms as authorities. Present the data, cite the source, move on.
0:08–0:14 — Where You Are (The Framework)
The AI Native Adoption Cycle — 3 minutes:
Hand out the printed diagram. Walk through quickly:
Stage 0 → Stage 1 → Stage 2 → Stage 3 → Stage 4 → Stage 5
UNAWARE CURIOUS EXPERIMENTING STANDARDIZING INTEGRATED AI-NATIVE
▲
MOST COMPANIES
ARE HERE — 95%
fail at this stage
“Most mid-market companies are at Stage 2. The 5% are at Stage 3-4. The gap is not budget. It’s organizational capability.”
Interactive — 1 minute:
“Quick show of hands — can anyone in this room tell me your organization’s total AI spend right now? All tools, all teams, including what employees are buying on personal accounts?”
[INTERNAL] Expect <10% of hands. When almost nobody raises a hand: “That’s normal. It’s also the first thing the 5% fixed.” This is the uncomfortable specificity moment — a gap about their own organization.
For the lawyers in the room — 2 minutes:
- Legal AI adoption doubled in one year: 23% to 52% of in-house teams (ACC)
- Stanford’s peer-reviewed study: legal AI tools hallucinate 17-34% of the time. Not a reason to stop — a reason to build verification workflows
- EU AI Act high-risk obligations take effect August 2, 2026. Colorado AI Act: June 2026. AI governance is shifting from best practice to compliance obligation
- 64% of in-house teams expect to depend less on outside counsel because of internal AI capabilities. That’s the competitive pressure.
[INTERNAL] This is why the workshop audience specifically cares. Their clients’ GCs are using AI tools that hallucinate 17-34% of the time. Frame as “solvable, not scary.”
0:14–0:22 — What to Do Monday Morning
The 90-day plan in 8 minutes:
Weeks 1-2: Audit
- Shadow AI audit — what are people already using, on what data? Takes 2 weeks. Typically reveals 3-5x expected spend.
- Take the 25-question AI Native Assessment. Know your stage.
Weeks 3-4: Decide
- Pick your tool based on your existing stack. If you’re a Microsoft shop, start with Copilot Business ($19/seat/month). It’s configuration — SSO, policies, license assignment.
- Write a 2-page AI acceptable use policy. Not 50 pages. Two.
- Budget honestly: the license is 4-9% of actual cost. Plan for training, governance, review process changes.
Weeks 5-8: Pilot
- 15-25 people. Mix of enthusiasts AND skeptics. Skeptics who convert become your best advocates.
- Train through champions (1 per 10-20 people), not mandatory webinars. 52% of employees use AI to game mandatory training rather than learn.
- Expect an adoption dip at weeks 3-5. This is normal. Don’t panic. Intervene with hands-on coaching.
Weeks 9-12: Evaluate
- Compile results into a business case with real numbers, not survey sentiment.
- Present to leadership: what we spent, what we measured, what we recommend for scale.
- The pilot either proved it or it didn’t. Either answer is valuable.
“The entire Tier 1 deployment is configuration. No custom development. No AI team required. The path to the 5% starts with doing week 1 well, not with buying more technology.”
[INTERNAL] Supporting: findings/90-day-quick-start.md. Hand out nothing during this section — send the full 90-day guide digitally after the session. It gives them a reason to open your follow-up email.
0:22–0:25 — Close
"AI works. The proof is UPS, JPMorgan, Citi, IKEA — sustained, measured returns over years. The 95% that fail don’t fail because of the technology. They fail because they buy tools without redesigning workflows, budget for licenses without planning for the real cost, and run pilots without deciding what success looks like in advance.
The framework on the page in front of you tells you where you are. The 90-day plan I’ll send after this session tells you what to do about it. If any of this raised questions specific to your organization, I’d welcome the conversation."
Leave the one-pager and adoption cycle printout on the table. Walk to the side of the room for informal conversation. Do not pack up.
[INTERNAL] The close is 90 seconds. No “in conclusion.” No recap of what you just said. No CTA beyond the natural “I’d welcome the conversation.” The follow-up email with the 90-day guide is the real CTA — it arrives when they’re at their desk thinking about it.
0:25–0:30 — Q&A
Likely questions and prepared responses:
| Question Pattern | Data Points to Reference |
|---|---|
| “Which tool should we buy?” | Don’t recommend a specific vendor. “Depends on your stack — I can walk through the decision framework after.” |
| “What about security/data leakage?” | 88% of orgs report AI security incidents. Assume Breach is the right model. Solvable with governance. |
| “How do we get our developers to use it?” | Champions, not mandates. Citi’s model. The adoption dip is normal — intervene at week 3-5. |
| “What about AI replacing jobs?” | IKEA reskilled 8,500, created $1.4B new revenue. Augmentation > replacement in every proven case. |
| “What does this cost, really?” | License fees are 10-20% of the real cost. For a 200-person company, $140K-$420K/yr on the people side alone (BCG 10-20-70). Budget 2.5x the license in Year 1 (DX Research/Atlan, 2025). |
| “Are we behind?” | “Most companies are at Stage 2. Being here and asking these questions puts you ahead of the 63% with no formalized AI initiative.” |
| “What about the legal risks?” | EU AI Act August 2026. Copyright unsettled. IP indemnification has loopholes. All manageable with governance. |
[INTERNAL] The Q&A is where deals start. Listen for the person who asks a question specific to their organization — that’s your follow-up conversation. Don’t answer with a sales pitch. Answer with data and “that’s worth a deeper conversation about your specific situation.”
Post-Workshop Follow-Up
Same day: Email to all attendees with:
- The 90-Day Quick Start guide (PDF)
- The AI Native Assessment link
- One sentence: “Appreciated the conversation today. The 90-day guide covers what we discussed — let me know if any of it raises questions for your team.”
Day 7: For anyone who took the assessment, send a brief personalized note referencing their likely stage. “Based on what you shared in the room, it sounds like your organization might be at Stage 2 — the 90-day guide’s first two weeks are designed exactly for that transition.”
Day 30: Brief check-in. “How did the first month go? Happy to compare notes.”
[INTERNAL] Do not send everything. Curate based on what the room asked about. If CFOs dominated the Q&A, lead with the Wasted Tokens Report. If security came up, lead with the Assume Breach analysis. Selective sharing of depth material builds the perception of a much larger knowledge base behind what they saw in 30 minutes.
Brandon Sneider | brandon@brandonsneider.com March 2026