Gradial home
Gradial and Reuters Events logos above the headline: AI is going to fix our marketing problems, right?
All blogs
InsightsNovember 17, 2026

How to move beyond AI experimentation to marketing transformation

Lynn Girotto, CMO, Gradial
Lynn GirottoCMO, Gradial
Insights

Congratulations, your marketing team has 47 new AI agents. Who’s managing them and how are they connected?

Yesterday, at the Reuters Strategic Marketing event, I opened my session with three questions for a room full of CMOs and marketing leaders.

  1. How many of you are already using AI somewhere in your marketing organization?

  2. How many of you feel confident explaining the ROI of marketing AI to your CFO?

  3. How many of you are sending AI agents across your organization without a clear way to coordinate and orchestrate all that work?

The last two questions are where things get interesting. One team is using AI to create content, another is building agents to automate campaigns, and someone else is experimenting with analytics. Who’s coordinating the handoffs, protecting the brand, and making sure these agents aren’t creating more chaos for everyone else? Bolting AI onto existing systems and processes won’t solve these problems. In fact, it only adds another layer of complexity to an already fragmented marketing organization.

Everyone is finding ways to move faster, but who’s making sure all that work is connected in a way that positively impacts the business?

These considerations bring us to the most fundamental question for marketing leaders: Are we using AI to experiment, or are we using AI to transform the business?

Experimentation isn’t an operating model

We’re all under pressure to grow the business, move faster, launch more campaigns, justify our budgets, and protect the brand while we’re at it.

So, AI is going to fix all of that for us. Right?

I realize that’s an interesting question coming from the CMO of an AI company. But nobody needs another lecture about how AI is changing marketing. We’re living it.

There’s always another release, another capability, another vendor promising the next big thing. Meanwhile, your team still has a campaign to launch on Thursday.

And if your CFO is shaking you down on the cost of running a multi-touch campaign with AI, good luck giving them a confident long-term answer. The technology is evolving, costs are changing, and the business case is still taking shape.

But the CFO is asking the right question. Are we getting a meaningful return on our investment?

We’ve all been prompting around and finding out. And experimentation is important. It’s how we learn what’s possible.

But experimentation isn’t an operating model. Making one task dramatically faster doesn’t transform marketing if everything around it still depends on disconnected systems, manual handoffs, and endless coordination.

In fact, you might just be creating a faster route into the same bottleneck. And the last thing we need is to re-silo marketing with a new generation of disconnected AI agents.

The cost of self-inflicted tech complexity

As marketers, we’ve invested heavily in technology designed to help us reach customers in more places, with greater relevance and personalization than ever before.

But somewhere along the way, we created a major problem for ourselves: We’ve invested in increasingly sophisticated technology, only to create an increasingly complicated way of working.

Think about where your team’s time actually goes: building pages, reformatting assets, moving content between systems, chasing approvals, and checking whether the final_final_final version is actually ready to launch.

Our most talented people spend an extraordinary amount of time operating the machinery of marketing instead of doing the work that creates value for customers.

Take a look at this resource triangle.

Today’s marketing effort: a large operational-execution base supports a smaller area for strategy and customer understanding. Execution includes creating, formatting, moving, editing, approving, launching, reporting, fixing, and coordinating.

At the bottom, where most of the effort goes today, sits operational execution. At the top, squeezed into a much smaller space, are strategy, creativity, and customer understanding.

AI gives us the opportunity to flip that triangle.

Imagine your team spending more time understanding customers, developing compelling ideas, improving experiences, and driving growth. All while the repetitive production and coordination work happens behind the scenes.

Future-state marketing effort: customer, brand, strategy, and creativity form the larger area, while operational execution becomes a smaller area powered by AI agents.

The operational work doesn’t magically disappear. It still needs to happen, and it needs to happen well. What changes is who, or what, does it. This is where AI can deliver value to the organization.

And we won’t get there simply by adding another AI tool to an already complicated stack. We have to rethink how the work moves across the entire marketing operation.

The three phases of marketing transformation

I like to explain the path forward with an analogy: the evolution of the car.

We’re not going to get from today’s marketing operations to an AI-native organization in one leap. But there are three phases that can get us there, and each one unlocks something fundamentally different.

Phase 1: Unlock the work

Think of today’s AI agents like driver-assist technology. Your car can parallel park, maintain speed, or help keep you in your lane. Individual tasks get easier, but you’re still driving the car. Marketing AI works much the same way today.

Agents can draft web content, update nurture programs, analyze feedback, build content variations, and check work against brand standards. That’s valuable. We’re giving people back time and improving productivity. But the underlying marketing operation hasn’t changed.

Someone still has to connect the dots, coordinate the work, manage approvals, and move everything through the same systems. This is where AI orchestration changes the equation. You need an agentic marketing harness to coordinate the workflows. Instead of deploying agents to perform isolated tasks, we connect them across workflows, with shared context, clear governance, and the infrastructure to execute at scale.

Think about everything required to launch a campaign: the brief, the assets, the build, the review, the approvals, the launch, and the optimization that follows. Today, people carry much of that work between systems. With orchestrated agents, the work can flow across those systems without someone manually managing every exchange. LLMs answer questions. Agents do the work. The agentic marketing harness brings it all together, coordinating execution across people and systems while maintaining brand consistency and governance.

LLM versus agent versus agentic harness: the LLM answers, the agent does the work, and the harness governs and coordinates execution.

That’s the shift Gradial is building toward: not just making individual marketers more productive, but enabling the entire marketing operation to work differently. And importantly, you don’t have to rip out your existing technology to get started. You can begin unlocking workflows in the systems you already have while building toward an infrastructure designed for agents.

And here’s the problem: as agents dramatically increase production capacity, they also create more work for the systems and processes downstream. Your agents can move at AI speed. Your legacy SaaS platforms can’t. Which brings us to the next phase.

Phase 2: Unlock the systems

You don’t get a self-driving car by adding one more feature to the dashboard. You have to rethink how the vehicle operates. The same is true for marketing. Our legacy SaaS systems were designed for people to operate them manually. They weren’t built for the speed, scale, or coordination required by thousands of agents. So why are we trying to force an entirely new way of working into systems designed decades ago?

In marketing, it means we have to look under the hood and rebuild the foundation. Why do we want to rebuild the foundation? Because we can do more. I recognize this feels daunting, and none of us wants to do a system migration. And let’s be honest—just thinking of a system migration is giving me PTSD. The number of website and CRM migrations that I’ve done over my career is time that I’ll never get back. And the sadder thing is the millions of dollars and months of work with zero value added for the business while it was happening.

This is why getting your marketing agents and AI harness in place first is critical. Once the agents are running the workflow across systems, the system just becomes a database. This makes the move to a native agentic system much faster and less costly.

The opportunity is enormous: unlock millions in legacy systems cost savings while avoiding the massive, time-consuming migration projects that have historically slowed marketing down. With agents and orchestration in place, you can modernize your infrastructure faster and move at the speed your business demands.

Phase 3: Unlock the potential

With your agents orchestrated and running on AI-native systems, you’re no longer constrained by the limitations of legacy technology. Now you can unlock what’s truly possible.

Now, imagine the car isn’t a car anymore. It’s a hovercraft. You’re no longer constrained by the roads someone built decades ago. You can get from point A to point B in ways that weren’t possible before.

That’s the opportunity ahead for an AI-native marketing organization. Instead of manually defining and coordinating every campaign step, imagine setting business objectives and allowing an intelligent marketing engine to determine, execute, and continuously improve the next best action based on customer signals and outcomes. An operation that learns and adapts. One that creates personalized experiences at a scale we couldn’t previously achieve.

We’re not there yet. But the decisions we make about our work and systems today will determine how ready we are when that future arrives.

Let’s be clear: marketers still set the destination. We decide which customers matter, what our brands stand for, what outcomes we’re pursuing, and where the guardrails belong. The goal isn’t to remove human judgment from marketing. It’s to remove the operational constraints that keep us from putting that judgment to better use.

You don’t have to wait for the hovercraft

We don’t have to be the Jetsons just yet. And the even better news is that you can unlock a lot of value in the first phase. Marketing transformation doesn’t require us to replace everything overnight. It starts with identifying where work gets stuck today and finding a better way to move it through the organization.

Two examples illustrate what’s already possible.

T-Mobile wanted to move faster and scale content production without sacrificing accuracy. Rather than simply adding more AI tools, they worked with Gradial to orchestrate execution across their existing systems.

The team reported 99% accuracy and an 80–90% improvement in time to market, without requiring a platform migration.

That’s an impressive operational result. But what interests me most is what happens when you give that capacity back to the people doing the work.

Amazon Music offers another perspective.

Its design organization produces more than a thousand assets each week. That’s an enormous volume of work, much of it repetitive and templatized.

When talented designers spend their days versioning assets and managing production, how much time is left to stretch their creative abilities?

By reducing repetitive work, the team has an opportunity to develop new skills, take on more complex creative challenges, and spend more energy on the craft they were hired to do.

“With just the little bit of automation…they’ve been able to learn new design tasks. They came in as static designers. They’re now learning motion design because that’s the way the business is going,” said Kelli Johnston, Sr. Design Ops Manager at Amazon Music.

That’s the human opportunity behind AI transformation.

Not simply getting more work out of the same people, but changing the work they’re able to do.

Time saved isn’t the win. What you do with it is.

Let’s say we successfully give our teams back a significant amount of capacity. What happens next? If we immediately fill that time with another pile of uncoordinated requests, we’ve made everyone busier without necessarily making marketing better.

But if we reinvest that capacity in the work that matters, the opportunity becomes much more interesting. We can get to market faster, tackling high-value initiatives that have been sitting in the backlog. We can do better work, giving teams room to test new ideas, improve customer experiences, and focus on the details that make a brand stand out. And we can develop new skills, helping talented people grow into the more strategic and creative work that AI makes possible.

For CMOs, this is where the business case becomes real. Saving time doesn’t automatically mean saving money. The value comes from what we do with that time.

Did we get a campaign to market faster? Improve an experience? Unlock an experiment we couldn’t previously support? Give our people room to take on higher-impact work? Those are outcomes we can measure and take back to the CFO. And they make a far more compelling case than simply reporting how many AI tools we’ve deployed.

Stop adding agents. Start redesigning the work.

Here’s what I’d encourage every marketing leader to consider.

Before you buy another AI tool or build another agent, ask yourself: What work is keeping my team from doing something more important?

Find a workflow that’s painful enough to matter and repeatable enough to improve. Understand where the handoffs break down, who owns the outcome, and what success should look like.

Then start there.

Prove the value. Learn from what changes. Use that experience to connect the next workflow and build toward a more orchestrated marketing operation.

Because the goal isn’t to prove that AI works. We already know it can.

The goal is to prove that marketing can work differently.

And build an organization where our best people spend less time operating the machinery and more time creating the ideas, experiences, and growth that move the business forward.

Download the event slides

Curious to see what AI orchestration could look like for your marketing organization? See Gradial in action.