The AI Application Layer - The Good, The Bad, And The Ugly | Jim Lecinski, Northwestern-Kellogg

A CMO Confidential Interview with Jim Lecinski, Clinical Professor of Marketing at the Kellogg School of Management, author, and former Google VP. Jim discusses why he believes marketers are often overly focused on using AI for productivity improvements versus business growth, the gaps between marketers and the C-Suite highlighted by recent Gartner research, and the difference between "big frontier models" and "shiny objects." Key topics include: why you should avoid "gray market AI", how to manage the 5 AI risks (privacy, accuracy, regulatory, personnel, and reputation), and the false precision that accompanies a focus on intermediate measures like Click Through Rate (CTR). Tune in to hear why he's not a fan of Cannes and how AI helped figure out a wedding invitation calling for "casual to semi-formal beach attire."

What should CMOs actually do with AI right now—and how do you avoid chasing shiny objects? Mike Linton sits down with Jim Lecinski, Professor of Marketing at Northwestern’s Kellogg School (and author of The AI Marketing Canvas and Winning the Zero Moment of Truth) to unpack the AI application layer: the good, the bad, and the ugly. Jim explains why CEOs-CFOs obsess over growth (not merely efficiency), how to reframe marketing dashboards around business outcomes, and his simple two-by-two for AI use cases (internal productivity vs. external value creation). We cover privacy, legal/regulatory, personnel, and reputational risks—and how to mitigate them—plus a pragmatic roadmap: center on a leading frontier model and layer vetted apps instead of stitching together fragile point solutions. Jim also shares candid takes on Cannes vs. Effies and ends with a challenge: personally build something with AI before year-end.

You’ll learn:

Growth over cost-cutting: aligning with CEO-CFO priorities and measuring ends, not means
The AI use-case 2×2: internal productivity vs. external, customer-facing value creation
Practical examples (e.g., apparel personalization) that lift CSAT, CLV, and revenue
The 5 risk buckets (privacy, accuracy, regulatory-IP, personnel, reputation) and guardrails
How to choose core models (GPT, Gemini, Claude) and avoid “tool soup”
Why awards that honor outcomes beat awards that celebrate activity

Guest: Jim Lecinski — Professor of Marketing, Northwestern Kellogg; former VP Customer Solutions (Americas) at Google; author of The AI Marketing Canvas (2nd ed.) & Winning the Zero Moment of Truth.

Host: Mike Linton — former CMO of Best Buy, eBay, Farmers Insurance; CRO of Ancestry.com.

Sponsor: Better marketing is built on Quad. See how better gets done at (https://www.quad.com/resources/resear...)

If you’re enjoying the show, please like, subscribe, and share with your leadership team. New episodes every Tuesday; companion newsletter on Fridays.