The AI Ops Person You Haven’t Hired Yet
Somewhere on your team, there is a person who has figured things out when it comes to AI.
They've built workflows nobody asked them to build. Or, they’ve quietly automated the task that used to eat three hours of your team’s time in meetings. Or, perhaps they know which tools talk to each other and which ones don't. And, they've stopped waiting for someone to tell them what's allowed—this is your AI operations marketing wizard.
Other people on the team have started bringing their questions to this person before they bring them to you.
Want more from The CMO?
Sign up for a free membership to complete reading this article:
Have an account? Log In
And while they may not have a title or even allocate dedicated time, they’re doing AI operations marketing that needs to be done. And you almost certainly haven't given them the organizational authority to apply what they know across the teams that need it most.
This is the gap. It’s not about the budget, or tooling, or even the strategy.
The Secret Sauce In Your Organization
This AI Operations role key to unlocking the potential with your organization, whether the details of the job are defined or not.
Nathan Snell, Director of Product Management at Intuit Mailchimp, has been building workflow automations for the better part of two decades. He has the data and insight to back up what works, as his workflows are stress tested against millions of users.
Snell refers to this person as an AI ops professional. He's been watching organizations succeed and fail at AI adoption. Long enough to know that this role, formally named or not, is the single biggest variable in whether an AI transition takes hold.
The variance in success is whether the person tasked with AI operations implementation has a mandate or just happens to be the most AI-curious one on the team. Perhaps they’re running experiments in their spare time with no organizational authority to change anything. Ultimately, it comes down to ownership.
And Snell’s diagnosis of where rollouts break is common.
"There are always two elements to any process," he says. "There's the mechanics of it, the work people are doing actively. And there's the coordination burden."
AI solves the mechanics faster than most organizations expect. But coordination, according to Snell, is something AI often doesn't touch. Without someone responsible for that gap, it can’t—and doesn't close.
What Lies Beneath
For marketing teams, the mechanics tend to accelerate between content production, data analysis, testing variations, and drafting. But speeding through tasks does nothing to improve the procedures and structures that can hinder organizations. Teams need to account for excess bureaucracy, processes, rules, and personnel that’s created when companies scale.
You might relate to the following when it comes to projects getting green-lit.
"If you have to have the pre-planning meeting and then the planning meeting and then the review meeting and then the get-it-live meeting and the how-it-went meeting," Snell says, "AI can't solve that. The coordination burden ends up having to be an actual process transformation."
Fractional CMO, and operations consultant, Colleen Goepfert marketing organizations on AI adoption, sees the same pattern from the outside. She's rarely brought in to fix what people say they want fixed.
"Workflow design matters more than the actual tool," she says. "The real work isn't designing the tool. It's redesigning how marketing works, how things get approved, how decisions get made."
What reinforces her point is how she describes the same dysfunction on the other side. Snell sees the impact of improved workflows in the products he builds and the customers who use them.
While Goepfert spots the obvious gaps in the organizations that were supposed to benefit from AI and haven’t yet seen that come to fruition. The conclusion is the same. AI doesn't create the problems within your team, cross-functionally or organizationally. But, it makes existing issues impossible to ignore.
"AI is really exposing a system that isn't working." - Colleen?
Akande Davis, VP of Operations at GNW Consulting, goes further. When his team enters an engagement, the first question isn't which tool to deploy. It's whether the breakdown is a process problem or a technology problem, because fixing the wrong one compounds the damage.
"Oftentimes technology can do above and beyond what the requirements are," Davis says. " [But] the process is broken." His analogy for what happens when organizations skip that diagnosis. “If I told AI to dig the hole in the wrong place, I get no value out of that hole. It'll just make things worse. You'll get worse results quicker."
The Architecture Most Teams Are Missing
Where Snell's perspective diverges from the diagnostic conversation is in what he recommends operationally. Most of the advice circulating about AI adoption centers on mindset, change management, and getting leadership aligned.
Snell's background as a product builder pushes him towards context and structure.
The value here cannot be understated, specifically how context is built, stored, and shared across teams. Most organizations approach this with a single large prompt or a shared document that gets outdated fast.
But Snell sees it another way. .
"Take this crazy prompt that you have, that was this massive gigantic thing in a template, and you can turn these into skills that are role-based. And then as a result, agents can actually do more. You can build on that context."
The shift from monolithic prompts to modular, role-based skills extends into operations.
And, it means the context an AI uses to do marketing work isn't the same context it uses to do product work or ops work, but all of it connects back to a shared foundation. Each team owns its layer. One AI operations person on a team maintains it.
"You really only need one person who can test the tools, create the skills, and share them," Snell says. "Centralizing it in that small way reduces friction versus having everybody go run experiments and then wondering why you're not seeing results."
Liza Adams, a fractional CMO and AI transformation advisor, describes the structural shift underneath this from a leadership angle.
"AI doesn't care about our silos. It doesn't care about who does the work or titles. It only cares about outcomes, which is good, because customers don't care about those things either."
She sees organizations moving, often reluctantly, away from hierarchies built around job titles toward structures organized around the job to be done. The AI ops person Snell describes is one manifestation of that shift: a role defined by outcome and workflow ownership, not by department.
The Part Coming Down The Pipe
What separates Snell's view most clearly from the rest of the conversation is that he's thinking past the current transformation. Everyone else is working to fix what's broken now. He's already designing for what comes after.
"I'm most excited about what I'd call the proactive agent transition," he says.
"Today, the vast majority of us experience AI in a pull-oriented way. I go to Claude or Gemini or ChatGPT. I ask it something, and I pull it out.
What we're going to see more of is push-oriented AI. For instance, if I've given you all my context, if you're connected to all my data, if you know how things are performing, why do I have to come to you and ask you to do something?
You should be able to look at it and say, 'I did this analysis for you, and here's a new segment of people you could be targeting. Do you want to use it?'"
For an ops leader, this is the reason the foundational work matters beyond this year’s AI goals. The teams building modular context architecture, where an AI pulls together the information it requires from different sources on demand are the ones defining who owns what. They’re also ensuring data isn't siloed across functions, which makes it possible to create the conditions for this kind of agent behavior to work.
The teams bolting AI onto existing processes are building a ceiling, one that’s co-dependant on individuals, making it both fragile and temporary.
Davis anchors the near-term version in how his team closes every engagement. Before recommending any change, they map back to value. Plotting where they can expect to see conversion lift, retention improvement, and pipeline contribution. And through this process they discover what’s vital for the organization.
"If it's not high value, let's focus on the things that are."
Operational redesign earns organizational support when someone is accountable for connecting it to outcomes the business already cares about.
Accountability is a human decision. And, no agent will be able to surface this until you've done the work to make it possible.
What’s Next?
Join us at The CMO Club for the latest insights, interviews, and podcast episodes featuring leaders in marketing.
