Keynote Speaker on AI and the Future of Marketing

What Your Team Will Learn

Six modules, each ending in applied work on your organization's real portfolio.

• The Fade, and Where It Stops. "Electric" dropped off "electric light" once the capability became assumed, and "AI-powered" is on the same path. The disappearance is uneven, because electricity is reliable and AI is probabilistic. Teams learn the qualifier-drop test and the cost-of-wrong grid that sorts every use case into the ones you let go quiet and the ones that stay named and watched.

• The Four Pillars: What AI Actually Does. Augmentation, Insights, Orchestration, and Generation replace the umbrella with four capabilities that behave differently, cost differently, and fail differently. Includes the author test, the boundary tool that separates Augmentation from Generation: remove the AI, and could a skilled person still produce this, just slower or at smaller scale?

• The Mechanism Axis: Nobody Brings a Forklift to Carry a Coffee Cup. Rules, predictive, generative, and agentic are four ways to build any of those capabilities, and the choice drives most of your bill and most of your risk. Teams learn the lightest-mechanism discipline: use the lightest mechanism that clears the task's accuracy bar. Most portfolios are paying agentic prices for work a five-line rule handled for free.

• The Supervision Dial. Autonomy isn't a category, it's a setting. Teams walk the full dial from human-approves-everything to human-never-looks, place their real workflows on it, and establish what has to be true before a system earns the next notch.

• The Four Layers You Can't Skip. Interface, Memory/Context, Trust/Verification, and Identity/Permissions are the conditions every capability requires and none of them performs. Includes the Trust-wears-four-faces exercise: what verification means under each pillar, and what the failure looks like when it's missing.

• Naming as Governance. The words a leadership team uses set the boundary of what it can think to ask. "Use AI responsibly" stops nothing, because there's no named act in it to permit or forbid. Teams write policy from real nouns instead, and build the on-call answer to the question that surfaces at 2 a.m.: what failed, and who do we wake up?

Every module lands in applied work. By the end of the session, the umbrella is off and there's something underneath it to manage.

The Three Instruments Your Team Builds

• The Vendor Decoder. A short interrogation you run on any "AI-powered" pitch. Which pillar are you selling me? Which mechanism is under it? What does it cost in tokens and in verification? What happens when it's wrong, and who finds out? A rep who can answer has decomposed their own product. A rep who dissolves every question back into "it's AI, it learns" has just told you the decomposition is about to become your job.

• The Portfolio Audit. Every line your organization currently calls "AI," sorted by pillar and mechanism, with the gap column that shows where you're running heavier machinery than the task requires. That column is usually the easiest savings anyone finds all year, because the over-built version works. It just works at many times the cost of the version that would have worked as well.

• The Fade / Stay-Named Line. A one-page test that sorts your use cases into what can safely become invisible infrastructure and what has to stay named, owned, and watched. Capabilities that work are exactly the ones that stop being monitored, and the monitoring was the whole job.

Who This Workshop Is For

This session is built for the people who have to answer for the spend:

• CMOs, CX Leaders, and COOs who need the AI line on the budget to survive contact with the CFO

• CEOs and Boards looking for a defensible read on what the organization has actually bought

• Marketing Operations and MarTech Leaders carrying a stack full of overlapping "AI" features nobody has decomposed

• Procurement and Vendor Management evaluating pitches where every product claims the same four letters

• Legal, Risk, and Governance Teams writing policy that has to cover autocomplete and an autonomous agent in the same document

• Cross-functional AI Councils and Steering Committees that have been meeting for a year without a shared vocabulary‍ ‍

The workshop works best when the room is mixed. The vocabulary problem is a coordination problem, and it resolves fastest when the people who disagree about what "AI" means are in the same room when the word comes off.

For Executives: What Your Organization Walks Away With

If you're a CMO, COO, CEO, or board member, here's what this session delivers.

You've approved the spend. What you may not have is a sentence you'd be comfortable defending to your CFO about what the money bought. This workshop produces that sentence, and the evidence underneath it.

What your leadership team brings back:

• A decomposed portfolio. Every "AI initiative" re-tagged by pillar and mechanism. Overlaps, gaps, and duplicate spend surface within the first hour of the audit, because they were always there and the umbrella was covering them.

• A named owner for every capability. Ownership attaches to a noun. "Who owns the generative content pipeline" has an answer with a name on it. "Who owns AI" has a committee and a recurring calendar invite that keeps getting moved.

• A cost story and a risk story that run independently. The mechanism gap column is your savings case. The cost-of-wrong column is your exposure. A capability can sit on exactly the right mechanism and still be dangerous, and the audit separates the two so you can act on each.

• Policy language that's actually enforceable. Rules written against named capabilities with named owners and named limits, rather than a values statement that governs nothing.

• Procurement leverage. The vendor decoder shifts the negotiation. You're no longer buying a category, you're buying a specified capability on a specified mechanism with a known failure mode.

• A realistic timeline for returns. The workshop covers the reorganization lag directly, using the historical case of factories that bought electric motors and bolted them onto the old driveshaft. The returns arrive when the work is rearranged, not when the technology is installed. Leadership teams that understand this stop killing programs one quarter before they would have paid.

‍ ‍

This is the session you book when "we should be doing more with AI" needs to become "here is what we own, what it costs, who runs it, and what happens when it's wrong."

Bring this session to your team →

For Practitioners: How to Make the Case Internally

If you're the person who already knows the portfolio is a mess, here's how to frame it upward.

Your leadership doesn't need convincing that AI matters. That argument is over. What's missing is a structured way to find out what the organization actually bought, and most of the available options are vendor demos wearing an education costume.

Talking points that work:

1. "It pays for itself in the audit." The gap analysis surfaces places we're running an agentic system where a predictive model or a rule would clear the bar. Same outcome, a fraction of the spend. That finding alone typically covers the cost of the session.

2. "It's our portfolio, not a case study." The working session runs on our real vendor list, our real stack, and our real budget slide. Nobody leaves with generic exercises.

3. "It answers the question legal keeps asking." We can't write an enforceable AI policy until we can name what we're governing. This session produces the nouns the policy needs.

4. "It's the vendor conversation we're about to have anyway." Every renewal this year will include a pitch with "AI-powered" in the first slide. The decoder gives us a way to price what's underneath it.

5. "It's led by Greg Kihlström." Host of the top-ranked enterprise marketing podcast, named the #1 Global Thought Leader on Customer Experience for three consecutive years, and author of the book this framework comes from.

6. "The outcome is measurable." Pre and post, we can count named owners, decomposed line items, and mechanism gaps closed. The audit is the artifact.