This article was based on the interview with From PegaWorld: Pega’s Tara DeZao on marketing ROI with agentic AI by Greg Kihlström, AI and MarTech keynote speaker for The Agile Brand with Greg Kihlström podcast. Listen to the original episode here:
We’ve all seen the demos. We’ve experimented with the tools. For the past two years, generative AI has been the talk of every marketing department, promising to revolutionize content creation and streamline workflows. Yet, for many enterprise leaders, the reality has been more complicated. Instead of a utopia of efficiency, we’ve often found ourselves managing a content assembly line that, while faster, can produce a glut of uncoordinated, off-brand assets. The operational bottlenecks haven’t disappeared; they’ve just moved. The challenge isn’t a lack of AI firepower, but a lack of intelligent orchestration. We have tools that can write, design, and code, but we’re still the ones manually connecting the dots from a strategic brief to a live, multi-channel campaign.
This is where the conversation among seasoned marketing leaders is shifting. The next evolution isn’t simply about creating more content faster. It’s about building an intelligent, autonomous system that can execute complex marketing tasks from end to end. This is the domain of agentic AI. It represents a fundamental shift from using AI as a discrete tool for a specific task to deploying it as an autonomous partner that can understand intent, manage workflows, and even provide strategic feedback. I recently sat down with Tara DeZao, Director of Product Marketing for Ad Tech and Martech at Pega, to discuss this very transition. Her insights underscore a critical turning point for marketing operations, moving from the chaos of creation to the clarity of governed execution.
Defining Agentic AI: The Connective Tissue for Marketing
Before we dive into the operational weeds, it’s crucial to establish a clear, no-nonsense definition of agentic AI. The term is gaining traction, but its meaning can get lost in the hype cycle. Unlike generative AI, which responds to a specific prompt to create an output, an agentic system is designed to take a goal, break it down into steps, and execute those steps autonomously. In marketing, this means bridging the chasm between a strategic document and a fully deployed campaign.
Tara DeZao frames it perfectly, moving beyond the idea of AI as a simple creator to AI as an operational hub. She sees it as the missing layer that translates intent into action.
“I consider agentic AI a layer… it’s the connective tissue between being able to create something with generative AI, let’s say, and then actually put it into production. So the agent acts autonomously on your behalf regardless of what it’s doing, whether it’s a marketing agent or a shopping agent… just to speed operations, make things faster and easier.”
This concept of “connective tissue” is vital for any marketing leader managing a sprawling MarTech stack. We don’t need another point solution. We need a system that can interact with our existing platforms—our DAM, our analytics suite, our personalization engine—and orchestrate their functions toward a common goal. An agentic framework isn’t about replacing these tools but about making them work in concert, guided by a central strategy. It’s the difference between having a team of individual specialists and having a project manager who ensures they all deliver on time and on brief.
From Brief to Campaign: Governance in Action
The true test of any new technology is its practical application. The promise of going from a campaign brief to a live, personalized experience in minutes instead of weeks is alluring, but as leaders, our first question is about control. How do we ensure this autonomy doesn’t lead to brand erosion or costly errors? This is where governance and human-in-the-loop oversight become non-negotiable.
Pega’s new Customer Engagement Studio is a tangible example of this philosophy. A marketer uploads a brief, and a suite of agents—a creative agent, a strategy agent, a compliance agent—begins to act on it. But this isn’t a black box. The system is designed to be a collaborator, asking clarifying questions and even flagging inconsistencies based on past campaigns. Tara highlights how this built-in validation provides a crucial safety net.
“If I am a marketer and I have uploaded a campaign brief to my agent, and I say, ‘I want to launch this over these channels,’…my agent might come back to me and say, ‘Okay, well, the last three briefs, you said to launch them over five channels, and now you’re telling me to launch them over four channels. Is that a mistake, or are you trying to do that?’”
This is a subtle but profound capability. It’s not just executing; it’s reasoning. It’s a pragmatic check against human error and a guardrail against what Tara calls “brand drift”—the slow, unintentional deviation from brand voice and strategy that can occur when autonomous systems are left unchecked for too long. For those of us in highly regulated industries like finance or healthcare, this level of embedded governance is essential. The goal isn’t to remove the human but to empower them with a co-pilot that catches the small things so the strategist can focus on the big picture.
Measuring What Matters: Adaptive AI and Customer Lifetime Value
For decades, marketing has been trapped in a cycle of measuring proxies for success. We chase clicks, opens, and engagement rates because they are immediate and easy to quantify. But we all know they don’t tell the whole story. The real measure of our success lies in sustainable, long-term customer relationships and lifetime value (CLV). The problem has always been connecting our day-to-day campaign activities to that long-term outcome.
This is where an adaptive AI model, working within an agentic framework, changes the game. It creates a continuous feedback loop where every customer interaction—or lack thereof—informs the next decision in real-time. A click is a signal, but so is a page dwell, a cart abandonment, or even the decision not to engage with an offer. Tara emphasizes that this ability to learn and pivot in the moment is what separates a truly intelligent system from a static, predictive model.
“That’s the most manual process of marketing, is, like, getting the results and turning around and making a change or a pivot. That takes the most time and is the most arduous thing, so that’s gonna be probably the biggest benefit to me, is being able to get your results, like, in the moment.”
By continuously adapting based on real behavior, the system can optimize for outcomes that matter, like CLV. As Tara notes, “marketers are often pinned down by the pressure of short term results,” but metrics like return on ad spend “are not going to tell you how you’re gonna do in five years.” An agentic, adaptive system allows us to keep an eye on both. It can optimize a campaign for this quarter’s numbers while simultaneously building a profile of customer behavior that informs a more valuable relationship over the next five years. It finally gives us the technical capability to align our short-term execution with our long-term strategic goals.
The Evolving Role of the Marketer: From Technician to Strategist
Naturally, the rise of autonomous systems brings up the inevitable question: what does this mean for our teams? If an AI agent can digest a brief, generate creative variations, configure a campaign, and optimize it in real-time, what is left for the human marketer to do? The answer, refreshingly, is the work we were all hired to do in the first place: think, strategize, and create.
The introduction of agentic AI doesn’t signal the end of the marketer; it signals the end of the marketer as a technician. The days spent manually configuring campaigns, pulling reports, and aggregating spreadsheets are numbered. Tara’s perspective is that this frees up human talent for higher-value work.
“Did we get into marketing to, like, aggregate a bunch of spreadsheets together? We did not. We came here to be creative… the strategic tasks are gonna still require humans and some of the more manual, okay, we need to configure, you know, and load up this piece of creative, those things are probably gonna fade a little bit.”
This evolution requires a shift in how we structure and skill our teams. The most valuable marketers in an agentic world will be those who can formulate a brilliant strategic brief, who can interpret the “why” behind the data, and who can provide the critical human oversight to ensure the AI’s actions align with the brand’s soul. The role of the marketing leader becomes less about managing tasks and more about defining the workflows and guardrails within which these powerful agents operate. We are moving from being players on the field to being the architects of the game itself.
The transition to an agentic marketing model is not an overnight flip of a switch. It is a strategic evolution that requires careful planning, robust governance, and a clear understanding of what we want to achieve. It’s about recognizing that not all problems require the same type of AI. As Tara aptly puts it, you don’t drive a Ferrari to the grocery store. Sometimes a simple predictive model is more effective and efficient than a complex generative one. The wisdom lies in using the right tool for the right job, and agentic AI is the master coordinator that ensures every tool is used to its full potential.
Ultimately, this new era of AI-driven marketing isn’t about removing humanity from the equation; it’s about elevating it. By automating the arduous and the repetitive, we create space for our teams to be more strategic, more creative, and more focused on the one thing technology can’t replicate: a genuine understanding of the customer. The future of marketing isn’t a world run by machines, but a world where machines empower humans to build more meaningful, valuable, and lasting connections with other humans. And that is something to be genuinely excited about.
