using-playbook

Using the AI Adoption Playbook

Purpose

Routes leaders to the right skill. Works for founders, CTOs, VPs of Engineering, CAIOs, COOs, consultants — anyone responsible for AI adoption. Enforces the rule that fluency-assessment runs first. If they've already completed the assessment, routes based on their request and scorecard results.

Flow

digraph router {
    "Leader arrives" [shape=box];
    "Mentions prior scores?" [shape=diamond];
    "Start fluency-assessment" [shape=box];
    "Has existing data?" [shape=diamond];
    "Bridge: map data to pillars" [shape=box];
    "Match request to skill" [shape=box];
    "Request unclear?" [shape=diamond];
    "Route by lowest pillar" [shape=box];
    "Invoke matched skill" [shape=doublecircle];

    "Leader arrives" -> "Mentions prior scores?";
    "Mentions prior scores?" -> "Match request to skill" [label="yes, has scorecard"];
    "Mentions prior scores?" -> "Start fluency-assessment" [label="no"];
    "Start fluency-assessment" -> "Has existing data?";
    "Has existing data?" -> "Bridge: map data to pillars" [label="yes — surveys, dashboards, etc."];
    "Has existing data?" -> "Invoke matched skill" [label="no — run assessment from scratch"];
    "Bridge: map data to pillars" -> "Invoke matched skill";
    "Match request to skill" -> "Request unclear?";
    "Request unclear?" -> "Route by lowest pillar" [label="yes"];
    "Request unclear?" -> "Invoke matched skill" [label="no"];
    "Route by lowest pillar" -> "Invoke matched skill";
}

Process

Step 1: Welcome and Start the Assessment

Do NOT ask "have you done an assessment before?" Most people haven't, and the question creates confusion. Instead, acknowledge what they're dealing with and go straight into the first question.

"Let's figure out where your team actually stands with AI. I've got a quick quiz — 9 questions, under 5 minutes. You pick A/B/C/D for each one, and at the end you'll have a scorecard with actual numbers across three areas: your team's mindset, how tools fit your workflow, and who owns making this work.

After that, I'd recommend sending a 5-minute survey to your team in parallel — your perspective is valuable, but hearing directly from the people using (or avoiding) the tools gives us a sharper picture.

First, two quick context questions. Tell me about your team — how big, what they do, and what's your role?"

This IS the fluency-assessment starting. You're not routing to it — you're beginning it directly. The survey is a parallel track, not a blocker.

If they mention having prior scores: Accept them and route forward. "Great — what were your scores? Even rough ones work."

If they mention having existing data (surveys, tool dashboards, engagement reports): Help them bridge it to our framework. See Step 1b.

Step 1b: Bridge Existing Data to the Framework

Many leaders have done employee surveys, have tool usage dashboards, or have some sense of where things stand — but not in our three-pillar format. Help them translate.

They have...	How to bridge it
Employee survey results	"What did people say about AI tools? Let's sort those responses into three buckets: resistance and fear, tool/workflow complaints, and 'nobody owns this' themes."
Tool dashboard / usage stats	"Usage data tells us about integration, but not psychology or ownership. Let's use your numbers for Pillar 2, and I'll ask a few questions to fill in the other two."
Anecdotal sense ("I think people are resistant")	"That's a starting hypothesis. Let me ask a few specific questions to confirm it and score it — your gut feel might be right, but the board needs numbers."
A previous assessment from a different framework	"Walk me through what you found. I'll map it to our three pillars — psychological barriers, integration failures, ownership gaps — and we'll see where the gaps are."

After bridging, you should have enough to produce a rough scorecard (even if some scores are estimated). Label estimated scores as "(estimated)" and proceed.

Step 2: Route Based on Request

Once a scorecard exists, match the founder's request to a skill:

Founder says something like...	Route to
"Board meeting coming up" / "need a board update" / "investor questions"	board-narrative-coach
"We're stuck" / "nobody's using it" / "adoption stalled"	blocker-diagnosis
"Where should we start?" / "what's the best first use case?"	first-use-case-picker
"Need a plan" / "roadmap" / "what do we do next?"	90-day-plan-builder
"Run the full process" / "start to finish" / "everything"	full-adoption-cycle
"How much are we spending?" / "are we wasting money?"	tool-stack-audit
"Show me the numbers" / "ROI" / "is it worth it?"	roi-calculator
"I need a snapshot" / "who's using what?"	adoption-scorecard
"Quarterly check-in" / "compare to last time"	quarterly-review

Step 3: Handle Unclear Requests

If the founder's request doesn't clearly match a skill, use their scorecard to recommend:

Lowest pillar	Recommend	Say
Psychological Barriers	blocker-diagnosis	"Your scores suggest people resistance is the main issue. Let's dig into what's blocking your team."
Integration Failures	first-use-case-picker	"Looks like the tools aren't fitting your workflow. Let's find the right starting point."
Ownership Gaps	90-day-plan-builder	"The biggest gap is that nobody owns this yet. Let's build a plan with owners and milestones."

Anti-Patterns

Skipping the Assessment

Symptom: Founder says "I know what I need, just help me with X." Consequence: Without a fluency scorecard, you're giving advice without diagnosis. Fix: "I hear you. But without understanding where your team actually stands, any advice I give is generic. The assessment takes 20 minutes and makes everything after it specific to your situation."

Routing to the Wrong Skill

Symptom: Founder says "we need a plan" but their real issue is that their senior engineers won't touch AI tools. Consequence: A 90-day plan built on top of undiagnosed psychological resistance will fail. Fix: If the scorecard shows a critical gap in one pillar but the founder wants a different skill, flag it. "Your scorecard shows a 1/5 on psychological barriers. A plan won't help until we understand that resistance. Want to run blocker-diagnosis first?"

Output

This skill does not produce its own artifact. Its output is invoking the correct next skill.

Next Skill

Whichever skill matches the routing logic above. Terminal state is the invoked skill taking over.

References

• fluency-assessment — always the first skill if no scorecard exists
• All other skills — this meta-skill routes to any of them