CXO Workshop · May 2026
A Strategic Playbook for Product & Technology Leaders
Today's Schedule
A structured half-day workshop designed to move from diagnosis to decision. Each module builds on the last.
Pre-Work Check
You should have arrived with one core customer workflow in mind: a process your customers perform repeatedly that your product facilitates.
You'll be mapping this throughout the session.
If you haven't done this yet, take 3 minutes now to jot it down.
AI inside the product, user-initiated. Copilot buttons. Summarize buttons.
AI as a premium add-on.
AI handles discrete tasks. Human reviews and approves.
Efficiency gains at the process level.
The product's core value proposition is delivered by AI. The workflow is the AI.
Human operates at exception level only.
Era 1: Feature Addition
Era 2: Workflow Automation
Era 3: AI-Native
What percentage of your AI investments are Era 1 (bolt-on features)?
Is there a single workflow in your product that AI fully owns end-to-end?
Would removing AI from your product change the core value proposition — or just a convenience feature?
We'll revisit your answers by the end of today.
"The companies that win in the next three years won't be the ones that added AI the fastest. They'll be the ones that figured out which workflows AI should fully own — and rebuilt their products around that."
The question for today: Which workflow is yours?
Workflow Deconstruction
Not every step in a workflow is equal. This module teaches you to see the difference — and find the leverage point.
Format
Framework walkthrough followed by a hands-on team exercise and group share-back.
Takeaway
A completed Workflow Deconstruction Map with seam points identified.
Duration
40 minutes total — 20 min framework, 20 min exercise and share-back.
| Classification | What it Means | When to Use it |
|---|---|---|
| Human-led, AI-supported | AI assists; human makes every decision | High-stakes, low-volume, regulatory exposure |
| AI-augmented | AI does the work; human reviews & approves | Quality gates matter; human judgment adds value |
| Agent-owned | AI executes end-to-end; human on exceptions only | High-volume, well-defined, reversible |
| Automation candidate | Rule-based; no AI needed — just eliminate | Workflow step has no intelligence requirement |
A seam point is where AI ownership would deliver a STEP-FUNCTION improvement in customer outcomes — not incremental efficiency.
Seam points = your highest-leverage AI investment decisions.
Key Question
"What would have to be true about our DATA, our MODELS, and our customer's TRUST for AI to own this step entirely?"
Exercise: Map Your Workflow
Take out your Workflow Deconstruction Map worksheet.
Write your customer workflow down the left column — 6 to 10 discrete steps.
Classify each step using the four categories from above.
Circle the seam points — steps where agent ownership would be transformational.
Star the one seam point you'd prioritize first.
AI-Native vs. AI-Added
Not every company can rebuild from scratch. But every company can find the one workflow worth rebuilding first. This module helps you tell the difference.
Format
Reframing session with the Lighthouse Project model, followed by structured group discussion.
Takeaway
A clear-eyed view of where AI-native thinking applies — and a candidate lighthouse opportunity.
Duration
35 minutes total — 20 min framework and cases, 15 min discussion.
Bolt-on AI adds intelligence to existing steps without changing how the workflow is structured.
Re-architecting around AI means the workflow itself changes — who does what, in what order, with what level of human involvement.
Both are valid. The question is: which steps in your customer workflow are candidates for re-architecture — and which are not?
Legacy code, acquired products, and short runway are real constraints. The methodology doesn't ignore them — it surfaces them at the right moment.
Start by identifying the highest-value workflows. Let feasibility constraints inform priority — not pre-empt the conversation.
A lighthouse project is one focused AI initiative — typically a single workflow or handoff point — designed to demonstrate what AI ownership looks like inside your specific product and organization.
It bypasses the "we can't rebuild everything" objection. You're not rebuilding everything. You're picking one thing. The goal is to show the team what's possible — and build the organizational muscle to scale it.
High-frequency, well-defined, reversible. Look for steps where AI can own the execution while humans stay on exceptions. The seam point you identified in Module 1 is your starting candidate.
The companies that succeed with AI don't start with a transformation. They start with a lighthouse. Then they ask: where's the next one?
Navigation simplifies. Users supervise rather than operate. The product surface shrinks as AI handles execution — which is a design challenge, not just an engineering one.
Seat-based pricing starts to break down when AI does the work. The most durable pricing model follows outcomes delivered, not access granted.
Features are replicable. Data loops are not. AI ownership creates a proprietary feedback loop — every decision improves the model that made it.
Product and engineering roles converge. AI systems need product judgment embedded in them — not bolted on after the fact in a sprint review.
The most valuable part of this session is what happens when CPOs and CTOs from different portfolio companies compare notes. Here is the prompt that will drive that conversation.
Discussion: Think about your portfolio company's most important customer workflow. Using what you've seen so far — where is the realistic lighthouse opportunity? Not the most ambitious re-architecture. The one you could actually prove out in 90 days. (5-min reflection → open discussion)
Listen for
Watch out for
Break
Reconvene for Module 3 — bring your Workflow Map
Use this time to review your Workflow Deconstruction Map and identify the seam point you'll bring into the Opportunity Matrix exercise.
The AI Opportunity Prioritization Matrix
Not all AI opportunities are created equal. This matrix cuts through the noise and forces the decisions that matter.
Format
Framework introduction, individual plotting exercise, then cross-portfolio peer review.
Takeaway
A prioritized AI opportunity map with peer-reviewed assumptions and three strategic overlays applied.
Duration
40 minutes total — 15 min framework, 15 min individual exercise, 10 min peer review.
Once you've plotted your opportunities, the map tells a story. Here's how to read each quadrant and make decisions from it.
Pilot
High value, uncertain feasibility.
Run a time-boxed experiment with a clear go/no-go trigger. Don't let uncertainty become inaction.
Invest Now
High value, high feasibility.
These are your 2026 bets. Assign ownership, define success metrics, and commit the resources.
Pass
Low value, low feasibility.
Explicitly park these. Write it down, make it visible, and protect the team's focus.
Research
High feasibility, lower value.
Don't build it just because you can. Feasibility without customer value is an engineering project, not a product bet.
↑ AI Feasibility
Customer Value →
New Revenue Stream Indicator
Does this create net-new revenue, or reduce cost / protect NRR? Mark opportunities that could generate new revenue streams with a star.
Build / Buy / Partner Decision
Does this require proprietary development? Or can it be assembled from existing AI infrastructure? Mark each opportunity.
Time-to-Value Estimate
Short-cycle (proof of value in 90 days) vs. long-cycle (12+ months). Mark each opportunity with S or L.
Exercise: Plot Your Opportunities
Phase 1 — 15 min: Individual
Plot your top 3–5 AI opportunities on your matrix worksheet. Apply all three overlays.
Phase 2 — 15 min: Peer Review
Swap with someone from a different portfolio company. Ask: "What assumption in your matrix would I push back on based on what I'm seeing in my market?"
Leadership Decision Points
Frameworks don't change companies. Decisions do. This module forces the specific choices that will define your AI trajectory.
Format
Role-specific decision frameworks for CPO and CTO, followed by an individual commitment exercise and group close.
Takeaway
A written 90-day mandate and a CPO/CTO accountability pairing before leaving the room.
Duration
45 minutes total — 20 min decision frameworks, 15 min commitment exercise, 10 min closing round.
Which customer workflow will AI fully own by end of year?
Not "what AI feature will we ship" — which workflow will be agent-led? Name it specifically.
What UX paradigm shift does that require?
How do you sequence the transition without breaking trust with existing customers?
How does this change your roadmap prioritization criteria?
The criteria you use with engineering must evolve when AI ownership is the goal, not feature delivery.
What is our data infrastructure position?
Do you have proprietary data loops that make AI defensible? Or are you training on the same data as your competitors?
What is our model strategy?
Build, fine-tune, or compose? And what's the trigger for re-evaluating?
How do you evaluate and govern AI outputs in production?
What is your trust architecture for agent-led workflows? Who is accountable when the agent is wrong?
The decision points in slides 4.1 and 4.2 are connected. The CPO cannot answer their questions without the CTO's input — and vice versa. Here is why that pairing is the unlock.
The companies executing AI-native transformation fastest are the ones where CPO and CTO have a joint AI product roadmap — not separate workstreams reconciled in sprint planning.
The AI architecture decisions are inseparable from the product strategy decisions.
What joint ownership looks like in practice:
A single AI roadmap document — not a product roadmap and a separate tech roadmap that get reconciled quarterly.
A shared definition of what "AI ownership" means for a given workflow — agreed before a line of code is written.
A weekly sync specifically on AI decisions — not buried in sprint planning or product reviews.
Explicit agreement on the trust architecture — who is accountable when an agent makes a wrong call in production.
Your Monday Morning Mandate
This isn't a reflection exercise. It's a commitment exercise. What you write here is what you bring back to your team on Monday.
"The one AI opportunity I'm committing to advance in the next 90 days."
Be specific: name the workflow, the customer segment, and what "advance" means — a decision made, a pilot scoped, a resource committed.
"The decision I've been avoiding that is the actual blocker."
Not a technical constraint — a leadership decision. Build/buy/partner? Which team owns it? What does "done" look like?
"One thing I learned from someone else in this room that changes how I'm thinking."
The peer perspective is the asset that disappears when you leave. Write it down now.
A note on the CPO/CTO pairing.
The companies executing AI-native transformation fastest share one thing: their CPO and CTO have a joint AI roadmap — not separate workstreams they reconcile in sprint planning. If you leave this room and pursue your AI opportunity alone, you will hit the ceiling that everyone hits. The architecture decisions are inseparable from the product strategy decisions. You have to solve them together.
Find your counterpart in this room before you leave. Share your Monday mandate. Agree on how you'll hold each other accountable.
Closing Round: 5 minutes
Each person reads their #1 commitment aloud — one sentence, no context needed.
Product School will follow up within 48 hours with a session synthesis, your commitment card, and optional office hours to pressure-test your next step.
Close & Playbook Review · 10 min
Each worksheet captures a different layer of your AI strategy. Together they form your 90-day action plan.
Your core customer journey with AI ownership classifications and seam points identified.
Top opportunities prioritized by customer value and AI feasibility, with overlays and peer feedback.
Your priority, blocker, and one insight from the room.
Product School will follow up within 48 hours with a session synthesis and optional office hours.