Alaura Weaver | September 17, 2026
In 2014, inside a room at Google, a group of executives watched a game of Pong. The paddles moved so fast you couldn’t track them, and the ball was a blur. Demis Hassabis explained that nobody programmed this. No rules, no if-then statements. The machine taught itself over a weekend.
Jim Lecinski was in that room. He’d spent more than thirty years in marketing, twelve of them at Google. He’d already helped rewrite the playbook on how customers make decisions with “The Zero Moment of Truth,” a framework that over 300,000 marketers downloaded and that ended up in textbooks. And watching that Pong demo, he had one thought: if this thing can learn to play Pong, maybe it could learn to recognize good customers from bad customers.
Two inflection points, one career. Jim was in the room when the web went commercial in 1994, and he was in the room when AI learned to play Pong twenty years later. Now a clinical professor at Northwestern’s Kellogg School, he’s spent thirty-seven years watching marketing get rewritten — twice.
- 85% of marketing teams are stuck in AI transformation. Only 15% of CEOs rate their marketing function as doing a good job with AI, according to Jim Lecinski. The problem isn’t technology — it’s treating AI adoption as a software purchase instead of a people challenge requiring change management, workflow redesign, and training.
- The Zero Moment of Truth has expanded from 3 sources to 5. Jim Lecinski’s original 2011 framework identified search results, ratings and reviews, and community conversations as the three sources influencing customer decisions. Generative AI added two new sources: content creators and answer engines (Perplexity, Gemini, AI search mode in Google). These sources are now interconnected — content creators feed AI engines, and community conversations become training data.
- The AI Marketing Canvas defines five stages of AI marketing maturity with no shortcuts. Jim Lecinski’s framework maps five stages: foundation (your own data), experimentation (structured, not “pilot-itis”), building capabilities (in-house, not borrowed), business case (real results for your CFO), and licensing (turning internal systems into revenue). No stage can be skipped.
- Generative AI added “generation” to the marketer’s toolkit, enabling “whole brain marketing.” For years, AI in marketing handled classification and prediction — the left brain work of finding patterns in data. The commercial launch of generative AI in November 2022 added the third capability: generation. Jim Lecinski calls this “whole brain marketing” — left brain for data, right brain for creation, both powered by AI.
- “Cognitive surrender” is the biggest risk for marketers using AI. Jim Lecinski warns against accepting AI’s first draft without questioning it. Every time marketers accept AI output without challenge, they chip away at their craft and their humanity. He advises marketers to prosecute and challenge AI output, maintaining their fundamental understanding of the craft.
From three sources to five: The new Zero Moment
Jim’s Zero Moment of Truth identified three sources that influenced customer decisions: search results, ratings and reviews, and community conversations. That was 2011. The framework changed how a generation of marketers approached their work.
But the map is getting redrawn. Jim is clear that the original three still matter. “Don’t ignore those three,” he says. “Racing towards something new and fun and different and shiny. You still gotta win those three.” But now there are five. The two new sources: content creators and answer engines — Perplexity, Gemini, AI search mode in Google.
The ecosystem is connected in a way it never was before. Content creators feed the AI engines. Reddit conversations become training data. Video transcripts shape the answers Perplexity gives your prospects. The whole system is talking to itself, and if you’re not part of the conversation, you’re not part of the answer.
Whole brain marketing
Jim’s explanation for why this moment feels different starts with what marketers have been doing all along. Marketing, he says, can be largely bucketed into three things: classification, prediction, and generation. For years, AI handled the first two — the left brain work of finding patterns in data and running the numbers. Then November 2022 happened. ChatGPT launched, and suddenly there was a third thing.
“With the commercial launch of generative AI, this now adds the ability for us as marketers to do the third thing with AI, that is the generation,” Jim says. “Now, as whole brain marketers, we have a whole brain tool moving forward.”
Left brain for data, right brain for creation, both powered by AI. Most marketing teams haven’t made the adjustment yet.
The AI Marketing Canvas: five stages, no shortcuts
Jim wrote his second book, “The AI Marketing Canvas,” before ChatGPT existed. He talked to hundreds of marketing teams and mapped what the winners did differently. It came down to five stages:
- Foundation — You need your own data. No data, no plan. But waiting years for the perfect data lake is also a bad plan.
- Experimentation — Not “pilot-itis,” where everyone with a corporate card does whatever they want. Structured experimentation with a plan.
- Building capabilities — You can’t build on borrowed land forever. Start taking some swings in-house.
- Business case — Real results that convince your CFO to invest. At this stage, you can make the cogent case for OpEx or CapEx.
- Licensing — A few companies will turn their internal system into a revenue stream.
“Newsflash for your listeners,” Jim says. “You can’t skip any stages.”
Why 85% of marketers are stuck
Five stages, straightforward and logical. So why are 85% of marketers stuck? Only 15% of CEOs rate their marketing function as doing a good job with AI.
Jim’s diagnosis: “The fallacy of thinking that this is just a technology software purchase licensing kind of a transformation where you buy something and do like a one-hour mandatory training and wish everybody good luck, as opposed to the human change management workflow changes, incentives, training, hiring, and promotion.”
Implementation isn’t a tech challenge. It’s a people challenge.
Freedom within a framework
Jim teaches jazz alongside marketing at Northwestern. When asked how the two connect, the answer came back to structure and freedom. People who haven’t been deep in jazz think it’s just a bunch of people playing what they want. But it’s the ultimate expression of freedom within a framework — a key signature, a tempo, chord progressions. The musicians come together, listen first, and create within that structure.
That’s AI for marketing. You set the parameters — the brand, the strategy, the guardrails — and then you create within them.
The warning: cognitive surrender
Jim had a warning that went with the freedom. “Don’t give into cognitive surrender,” he says. “Cognition is one of the higher order attributes that sets humans apart.”
It’s tempting. We all have a to-do list an arm and a half’s length long. Just do this for me and then accept whatever comes back. “Every time we do that, that sort of chips away at a little piece of our collective and our individual humanity.”
Prosecute it, challenge it. Don’t just accept. The craft still matters, maybe more than ever. “The fact that ChatGPT can answer a question does not exclude or excuse the need to have a fundamental understanding of the craft yourself.”
What we’re building toward
AI is squarely at the center of the plate of what marketers do. Not a nice-to-have, not a pilot program. The main thing. But implementation is the people work.
So what are we building toward? A world where left brain and right brain work together, where the customer is always in the room — even synthetically — where we win in five places instead of three, where freedom lives inside frameworks, and where the craft still matters. And where cognitive surrender is the one failure mode we refuse to accept.
Your Monday morning action: pick the workflow where you’ve stopped editing the AI’s first draft. This week, prosecute it, challenge it, and see what improves when you stop surrendering.
Framework: The AI Marketing Canvas
Jim Lecinski developed the AI Marketing Canvas after talking to hundreds of marketing teams about their AI adoption efforts. He asked each team one question: are you having success, making good progress in your own estimation, or is this not working? Then he mapped what the winners did differently.
The framework identifies five stages of AI marketing maturity:
Stage 1: Foundation
Build your data infrastructure. You need your own data — but don’t wait years for the perfect data lake. “If you try to charge ahead with AI with no data, that’s not a good plan,” Jim says. “On the other hand, if you say ‘we’re waiting seven years for our management consulting partner to be done with the perfect, amazing data lake’ — bad plan.”
Stage 2: Experimentation
Run structured experiments — not uncontrolled pilots. Jim warns against “pilot-itis,” where everyone with a corporate card does whatever they want in a free-for-all with no plan. The best practice is deliberate, governed experimentation.
Stage 3: Building Capabilities
Develop in-house AI skills. “You can’t build on borrowed land forever, as the saying goes,” Jim says. “You have to start taking some in-house, some swings. Making some bets on your own.” This is where teams move from relying on external tools and consultants to building their own AI capabilities.
Stage 4: Business Case
Document real results and build the case for investment. “At stage four, you can go back to your CFO and you’ve got the ability to build a business case and say, here’s the lift from when we apply AI,” Jim says. You may need OpEx, CapEx, or new hires — but at this stage, you can make the cogent case to your CFO and board.
Stage 5: Licensing
A few companies will turn their internal AI system into a revenue stream by licensing it externally. “Stage five, just a few companies will get there,” Jim says. “That system which they have built for their own internal use, perhaps they can license as a revenue stream externally.”
The critical rule: you can’t skip any stages.
FAQ
How do enterprise marketing teams move from AI pilots to production?
Enterprise marketing teams move from pilots to production by following a staged approach. Jim Lecinski’s AI Marketing Canvas identifies five stages — foundation, experimentation, building capabilities, business case, and licensing — and none can be skipped. The most common mistake is treating AI adoption as a technology purchase rather than a people challenge that requires change management, workflow redesign, training, and incentives.
What are the five stages of the AI Marketing Canvas?
The AI Marketing Canvas outlines five stages of AI marketing maturity:
(1) Foundation — building your data infrastructure(2) Experimentation — structured testing, not uncontrolled “pilot-itis,” (3) Building capabilities — developing in-house AI skills rather than relying on borrowed land(4) Business case — documenting real results to secure CFO investment(5) Licensing — turning internal systems into revenue streams. Jim Lecinski emphasizes that no stage can be skipped.
How has the Zero Moment of Truth changed with generative AI?
The Zero Moment of Truth has expanded from three influence sources to five. The original three — search results, ratings and reviews, and community conversations — still matter. Generative AI added two new sources: content creators and answer engines (Perplexity, Gemini, AI search mode in Google). These sources are now interconnected, with content creators feeding AI engines and community conversations becoming training data.
What is cognitive surrender and why should marketers avoid it?
Cognitive surrender is the tendency to accept AI-generated output without questioning or editing it. Jim Lecinski warns that every time marketers accept AI’s first draft without challenge, they chip away at their craft and their humanity. He advises marketers to prosecute and challenge AI output, maintaining their fundamental understanding of the craft even as AI becomes more capable.
How can marketing teams use synthetic personas?
Synthetic personas are AI-built customer personas that keep the voice of the customer present between real research touchpoints. Rather than replacing focus groups or tracking studies, they supplement them by providing a customer perspective during day-to-day decisions. Jim Lecinski emphasizes that synthetic personas should not replace real customer conversations but can keep the customer “in the room” when making decisions on a Tuesday afternoon.
Why is AI implementation a people challenge rather than a technology challenge?
AI implementation fails not because of technology limitations but because of inadequate change management. Jim Lecinski found that successful AI marketing transformation requires workflow changes, incentives, training, hiring, and promotion adjustments — not just a software purchase and a one-hour training. Only 15% of CEOs rate their marketing function as doing a good job with AI, largely because teams treat it as a tech problem rather than a people problem.
