At A Glance
Agentic AI is moving faster through media planning than most organizations can define what it means for their business. Vendors are pitching it, conferences are built around it, and marketing teams feel pressure to move now. Before automation reallocates budget or reshapes a customer journey, marketers need three things in place: a shared definition of agentic AI, an identity foundation the system can trust, and independent measurement that keeps buying and performance separate.Agentic media planning has gone from emerging idea to assumed industry direction in a single cycle. The instinct to move fast on new technology makes sense. When a whole industry agrees on something quickly, though, the definitions, limits, and trade-offs that matter most tend to get skipped, and “agentic AI” still means very different things to different teams. Organizations building toward it without a shared definition, risk creating incompatible systems under the same label. There’s foundational work to do before automation takes the wheel.
What does agentic AI mean in media?
Much of the current enthusiasm around agentic AI conflates the technology with a specific tool. In many conversations, “agentic” has become shorthand for highly autonomous generative AI, as though large language models (LLMs) were the entire system rather than one component within it.
That framing understates what agentic systems are. Agentic systems are decision frameworks, not models. They coordinate predictive models, generative tools, and non-AI logic to move from signal to recommendation to action.
This pattern already exists outside media. Starbucks’ “Deep Brew” connects real-time purchase signals with personalized offers while informing what products stores stock and promote. The system coordinates customer demand with operational decisions across the business. Layering prompt engineering onto an LLM is a different thing entirely.
Without a shared definition of what “agentic” means for your organization, teams risk optimizing different visions of the same idea:
All may qualify as agentic, but without a shared definition, you can’t build toward the same goal. That’s why defining the problem statement comes first, not last.
Why do acting systems raise the stakes?
Here’s what sets agentic systems apart from earlier waves of automation: they don’t just inform decisions; they execute them. Programmatic has always operated within guardrails, executing and optimizing based on defined parameters. Agentic systems work more like navigation. They determine direction and weigh trade-offs against a wider set of context. When an agent reallocates budget mid-flight, suppresses an audience segment, or swaps creative, the quality of the identity and business logic behind that decision determines the outcome.
Here’s what sets agentic systems apart from earlier waves of automation: they don’t just inform decisions; they execute them.
Apply that framework to media. An agentic system operating on incomplete identity data or stale signals won’t slow down to account for the gap. It optimizes faster in the wrong direction. The coordination layer needs to weigh reach, incrementality, audience fatigue, and margin before it shifts spend, and that requires inputs that are accurate, current, and governed. Accountability and oversight aren’t optional anymore; they need to be built into the design from the start.
Watch our Curated Couch session on agentic media planning
Why should measurement stay separate from buying logic?
As autonomy grows in media systems, a structural problem emerges: the system shouldn’t measure its own performance.
When buying logic and measurement logic live in the same loop, performance numbers can look strong simply because they’re optimized to the system’s own signals. Success stops being defined externally and starts being defined internally. Agentic models trained on this pattern learn from the signals they favor, optimizing toward the outcomes they’re best equipped to observe. Performance looks good inside the system, and reflects a narrower view of what’s actually happening.
Independent measurement gets framed as friction on automation more often than not. In reality, it’s what lets automation to scale responsibly. Incorporating independent measurement signals into agentic decision-making, while maintaining meaningful separation from buying logic, produces an objective view of performance, one that can be compared consistently across systems rather than optimized within a single one. That distinction matters when you need to explain results to a client, stakeholder, or regulator.
AI as coordination layer, not an infrastructure replacement
Nothing here slows down automation. Independent measurement doesn’t add friction to execution but builds confidence in the outcomes those decisions generate. The organizations that will benefit most from agentic media won’t necessarily be the ones that adopt it fastest, but those that take the time to define where trusted autonomy creates measurable value and where human judgment needs to stay in the loop.
Before automation reallocates dollars or reshapes customer journeys, the data model has to hold up under scrutiny.
Trust in data, systems, and decision-making processes becomes the real constraint as agentic systems move from insight to action. That trust doesn’t come from autonomy alone, but it comes from governed data and decisions that can be explained, audited, and defended. Every brand that has operationalized automation at scale has learned the same lesson: before automation reallocates dollars or reshapes customer journeys, the data model has to hold up under scrutiny.
How does Experian help you build a defensible foundation?
Experian’s approach to AI in marketing starts from the same principle: intelligence is only as reliable as the foundation it’s built on.
We have spent decades building and governing data in regulated industries. That experience is what makes our foundation AI-ready: not just technically capable but built to withstand the scrutiny that responsible automation requires.
2026 State of advertising report
Our 2026 State of advertising report brings together perspectives from 14 leaders operating across key parts of the advertising ecosystem to show how shifts in AI, commerce media, healthcare, and more, are taking shape in practice.
Download our 2026 State of advertising report and hear from Scott Bender, Head of Publisher & Platform Partnerships, Newton Research, on how AI in advertising is reshaping planning, activation, and measurement.
The data foundation comes first
Agentic AI will reshape how media is planned and bought. That shift is happening now, and the opportunity is significant. But the advantage belongs to organizations that build on a foundation their systems, partners, and stakeholders can trust. Before automation runs, the inputs must be defensible, identity must be accurate, measurement must be independent, and the problem statement must be clear. Get those things right, and automation becomes more powerful. Skip them, and you’re simply moving faster in the wrong direction.
About the author
Budi Tanzi
SVP, Product, Experian
Budi Tanzi is the Senior Vice President of Product at Experian Marketing Services, overseeing all identity products. Prior to joining Experian, Budi worked at various stakeholders of the ad-tech ecosystem, such as Tapad, Sizmek, and StrikeAd. During his career, he held leadership roles in both Product Management and Solution Engineering. Budi has been living in New York for almost 11 years and enjoys being outdoors as well as sailing around NYC whenever possible.
FAQs
Agentic AI refers to systems that move from signal to recommendation to action with a high degree of autonomy. In media, this means an agent can reallocate budget, adjust audience targeting, or swap creative mid-flight based on real-time signals. Unlike standard programmatic systems that execute within predefined guardrails, agentic systems function more like a navigation layer, coordinating across data inputs, predictive models, and business rules to make directional decisions.
Data quality matters more as media automation becomes more autonomous because autonomous systems don’t slow down to account for bad inputs. If an agentic media system operates on incomplete identity data or stale audience signals, it optimizes faster toward the wrong outcomes. The quality of the identity foundation, the accuracy of audience attributes, and the reliability of measurement inputs all become more consequential, not less, as human oversight is reduced.
Programmatic executes and optimizes within defined parameters whereas agentic systems function at a higher level of abstraction, weighing reach, incrementality, fatigue, and margin before determining where and how to shift spend. Programmatic is the engine. Agentic AI, when built correctly, is the navigation layer that determines direction based on wider context and business objectives.
Measurement should be kept separate from buying logic in agentic systems because when a system measures its own performance, it optimizes toward outcomes it’s best equipped to observe. The definition of success becomes internal rather than external, and performance numbers may look strong while reflecting a narrower view of what’s working. Independent measurement, kept meaningfully separate from buying logic, produces results that can be compared across systems and explained to stakeholders without ambiguity.
Experian’s Offline Graph and Digital Graph provide a privacy-first identity foundation covering more than 250 million U.S. consumers and 4.2 billion digital IDs. Experian’s Marketing Data adds more than 5,000 attributes per audience, and Experian’s outcomes measurement keeps performance signals independent from buying logic.
Latest posts
Commerce media networks have had a strong start. Growth has been fast, demand has been strong, and brands have made it clear they want closer access to commerce-driven audiences. But as more networks mature and enter the space, many are starting to feel the same pressure point: scale. Most commerce media networks were built as managed service businesses. That model works well early on. High-touch, white-glove partnerships make sense when you’re working with a handful of strategic brands. But there’s a ceiling. There are only so many teams, only so much inventory, and only so many advertisers that model can realistically support. It’s one thing for a large retailer to build custom programs for a P&G. It’s another to do that at scale for hundreds or thousands of brands. At some point, growth slows, not because demand disappears, but because the model can’t stretch any further. The scale problem no one likes to talk about That’s where many commerce media leaders find themselves today. Pausing to assess what comes next. For a long time, growth has been measured almost entirely through media dollars. That mindset is understandable. Media is familiar, it’s easy to quantify. It shows up clearly in negotiations and revenue reports. But viewing commerce media networks purely as media sales engines creates long-term risk. It can strain brand relationships, limit innovation, and distract from what commerce media networks actually do better than almost anyone else: understand consumers deeply. Signals are the real asset Commerce platforms sit close to decision-making. They see what people search for, what they consider, what they buy, and when those behaviors change. Those signals are incredibly powerful. And yet, most networks only activate them inside their own walled environments. That’s a missed opportunity. Curation represents the next area of growth for commerce media networks, and it doesn’t require replacing or diminishing existing media revenue. In fact, it complements it. No single commerce media network has all the data needed to give advertisers the scale and reach they’re looking for. And no advertiser wants to recreate the same audience in dozens of disconnected platforms. That friction creates inefficiency and slows decision-making. Why collaboration supports sustainable growth The opportunity is to look beyond first-party data alone and start thinking about collaboration. Second-party data. Data partnerships. Signal sharing done responsibly and transparently. Imagine an advertiser defining an audience once and being able to understand and reach that audience across multiple commerce environments. Not through a series of disconnected buys, but through a more consistent approach built on shared understanding leading to increased reach and more impactful campaigns. That’s easier for advertisers to manage, and it creates an additional revenue stream for commerce media networks that complements media sales rather than competing with them. Curation strengthens media, it doesn’t replace it Media will always play an important role. There is clear value in custom experiences tied directly to a commerce environment. Think buyouts, sponsored experiences, custom creative integrations. Those are situations where brands want to work closely with the network itself. But the signals commerce media networks hold don’t need to be limited to those moments. Those signals can be monetized independently through data products, co-ops, and partnerships that extend their value into other channels. That’s how curation adds value without undercutting existing revenue. A practical path forward for commerce media leaders For commerce media leaders thinking about their next phase of growth, the focus should be on sustainability. Building a massive media operation takes time and investment. Data-driven revenue streams can be introduced more quickly, require fewer internal resources, and provide steadier margins. It’s a practical approach. Use signal-based revenue to fund growth. Let that revenue support investment in tooling, talent, and media innovation over time. Bootstrapping, in the truest sense. Why transparency matters early There’s also a broader responsibility here. In many advertising channels, transparency followed growth, often after pressure from the market. Commerce media networks have an opportunity to do this differently. To lead with transparency from the start. To be clear with brands and consumers about how data is used, how signals are created, and how value flows through the ecosystem. Because the reality is this: commerce media networks are holding some of the most valuable intent signals in the market today. But those signals don’t retain their value in isolation. If they aren’t enhanced, combined, and made accessible in the right ways, someone else will step in to do it. And when that happens, control shifts away from the source. The bottom line The next chapter of commerce media isn’t just about selling more media alone. It’s about recognizing the value of the signals already in hand, working together to make them more useful, and building additional revenue streams that support long-term growth. That’s how commerce media networks grow without eating their own lunch. About the author Kevin Dunn Chief Revenue Officer, Experian Kevin Dunn joins Experian Marketing Services with more than 20 years of leadership experience across marketing and advertising technology, most recently serving as Senior Vice President of Brands and Agencies at LiveRamp. In that role, he led growth across retail, CPG, travel, hospitality, financial services, and healthcare, overseeing new business, account expansion, and channel partnerships. Kevin is known for building cohesive, accountable teams and leading with optimism, clarity, and a strong sense of shared purpose. His leadership philosophy centers on empowering people, driving positive outcomes for clients and fostering a culture where teams can grow, take smart risks, and succeed together. Latest posts
Learn why programmatic curation is becoming the standard for privacy-first, performance-driven media buying in 2026.
Learn how energy and utility marketers use Experian Audiences to reach households based on energy usage, sustainability interest, and tech adoption.