Quarterly review with your CMO. She walks in prepared. The deck is good. The numbers from last quarter are mostly green. Pipeline is steady, brand metrics moved a little, the team morale slide has the usual photos.
You sit there listening, and somewhere around slide twelve you feel something is off. You can’t name it. You ask a couple of questions. She answers them well. You sign off on the quarter and tell her good work and the meeting ends on time.
Then for the next two days you keep thinking about that feeling. The deck was good. The answers were fine. So what was off?
Here is what was off. Your CMO is running marketing the way she ran it in 2022 and she has not noticed that the job description has changed. The deck is good because she is good at the deck she’s been giving for the last four years. The board scorecard your investors started using six months ago is asking different questions, and her deck doesn’t answer any of them. And nobody has told her that, because the questions show up as instincts in the room before they show up as agenda items.
She is failing a test nobody told her she was taking. You are the only person who can either tell her or replace her, and most CEOs end up doing the second because they didn’t know how to do the first.
What the Numbers Say
Gartner published a survey result in February 2026 that should have triggered an executive conversation in every B2B leadership team. 65% of CMOs expect AI to fundamentally alter their roles. Only 32% believe significant personal skill changes are needed.
Two-thirds of marketing leaders see the wave coming. Less than a third think they personally need to do anything different about it.
This isn’t a knowledge gap. The data is everywhere. Every analyst, every consultancy, every conference keynote has been hammering the AI transformation message for two years. CMOs know the story. They can recite it on cue. They quote it in their board updates and the slides land well.
What they cannot do is apply it to themselves.
The disconnect goes deeper. Only 15% of CEOs said they consider their current CMO to be AI-savvy. 85% of CEOs are sitting across the table from a marketing leader they do not believe is ready for what is coming. By 2027, Gartner predicts that lack of AI literacy will be a top-three reason large enterprise CMOs are replaced.
If you are in that 85%, the feeling you couldn’t name in the quarterly review is the early version of a board conversation that is going to happen with or without your input.
The Feeling Before the Conversation
When CEOs say they want an AI-savvy CMO, most CMOs hear „I need to use ChatGPT more.” So they start using ChatGPT more. They generate a few campaign briefs with it. They write a LinkedIn post about how transformative the tool is. They check the box and assume the conversation is over.
Sure, that is one use of the tool. But it isn’t what AI-savvy means in 2026.
AI-savvy means knowing which workflows in your marketing organization should be rebuilt around model output and which should stay human-led. It means understanding what model your team is using, what its limitations are, and what your data exposure looks like when an analyst pastes a customer list into a public chatbot. It means having an opinion on whether the agency you just hired is using AI to scale a senior strategist or to replace junior writers with model output the client is paying senior rates for.
It means understanding the difference between AI as a productivity tool and AI as an operational layer that changes how the whole department gets its work done. The first is a tactical change. The second is an organizational redesign.
CMOs who treat the second like the first are the ones the board is talking about replacing. Forrester’s 2026 B2B Predictions report estimates that B2B companies will lose more than $10 billion in enterprise value due to ungoverned use of generative AI: declining stock prices, legal settlements, fines, brand damage. The losses will concentrate in companies whose marketing leaders saw AI as a tool to accelerate their existing playbook instead of as a discipline that needed leadership-level governance.
If your CMO can’t explain to you, in two minutes, what her team’s exposure to ungoverned AI looks like and what she’s doing to control it within her function, that’s not a tactical gap. That’s the gap the board is going to use to decide whether she’s the right person for the next stage.
What the Board Is Actually Looking For
I’ve been on enough calls with boards evaluating marketing leaders to know what they look for, even when they don’t say it out loud. They look for three things.
First, evidence the CMO has redesigned at least one major workflow because of AI. Not „we are exploring AI options.” Actual changes. The content review process is different now. The campaign briefing process is different. The vendor evaluation rubric is different. Something visible has changed in how the marketing organization gets its work done. If the CMO says „we use AI for content drafts now” and that’s the entire change, the board knows the answer. The change is cosmetic.
Second, evidence the CMO has an opinion on governance. What is the policy on AI tools? What data can the team paste into public models? What gets reviewed by humans before it ships? What happens when an AI-generated piece of content gets flagged as inaccurate? CMOs who can’t answer these questions are signaling that they aren’t paying attention at the level the role requires.
Third, evidence the CMO is reshaping the team. Not necessarily smaller. Different. New roles, new skills, new accountability structures.
The marketing org chart from 2022 isn’t the right chart for 2026, and the CMO who hasn’t redrawn it is the CMO who hasn’t noticed the change.
These aren’t technical requirements. They are leadership requirements. They are exactly what good CMOs have always done when their function was being reshaped by external forces. The difference is that AI is reshaping the function faster than most CMOs have ever experienced, and the board’s patience is shorter than the CMO’s learning curve.
The Comfort Trap
I see this most clearly in fractional engagements with companies whose CMO has a strong track record from the previous era. The CMOs who are most resistant to retooling their own thinking are almost always the ones with the strongest 2020-2023 results. They built marketing teams that worked. They hit revenue targets for years. They have a network of agency partners they trust and a content playbook that produces predictable output.
The instinct is to apply AI as a layer on top of the playbook that made them successful. Faster content production. Smarter targeting. Better attribution. The playbook stays. The tools improve.
The problem is that AI doesn’t just speed up the playbook. It changes what the playbook should be.
The team structure that produced results in 2023 may not be the right structure for 2026. The agency that made sense when content production was the bottleneck may not make sense when the bottleneck is editorial judgment over a flood of AI output. The metrics that defined success when humans wrote everything may not capture what success looks like when humans review and direct.
The CMOs who got promoted on the old playbook are emotionally invested in the playbook. They built it. They defended it. It made them successful. Telling them the playbook is now the wrong playbook feels like an attack on their judgment, not an update to the conditions.
That comfort trap is also the firing trap. The CEO is having the conversation the CMO isn’t having with herself.
Three Things to Watch For at the Next Quarterly Review
You don’t need to fire your CMO this quarter. You need to know whether the conversation is overdue. Walk into the next quarterly review and watch for three signals.
When she presents the marketing operations slide, does she describe a workflow that has visibly changed because of AI, or does she describe the same workflow with AI tools bolted onto the side? The answer will tell you whether she sees AI as an operational layer or as a productivity tool.
When you ask her what her team’s exposure to ungoverned AI looks like, can she give you a coherent two-minute answer that names the risks, the policies, and what’s reviewed by humans before it ships? If the answer is fluent, she’s been thinking about this. If the answer is hand-wavy, she hasn’t.
When you ask her how the marketing org chart should look in 18 months, does she describe new roles and new accountability structures, or does she describe the same chart with the same titles? The CMO who can’t draw a different chart is the CMO who is going to be replaced by someone who can.
If two out of three signals come back wrong, you have a conversation to have. Not a firing conversation. A conversation about what specifically you’re going to ask her to retool, on what timeline, and how you’ll know whether the retool is working. The CMOs who make it through this transition will be the ones whose CEO had that conversation with them six months before the board started asking questions. The CMOs who don’t will be the ones whose CEO waited for the board to ask first.
You’re the early warning system. The board scorecard your CMO can’t see runs through your office. If you don’t tell her the test exists, nobody will, and she’ll fail it without ever knowing she was being graded.
